Endoscopy in the Intensive Care Unit

  • Harvey M. Licht
  • Fredric Jaffe
  • Gilbert E. D’Alonzo


After studying this chapter, you should be able to do the following: Understand the indications for both respiratory and gastrointestinal (GI) endoscopy in critically ill patients. Describe the contraindications and potential complications that are associated with endoscopy and endosurgery.


Bronchoalveolar Lavage Endotracheal Tube Esophageal Varix Blunt Chest Trauma Fiberoptic Bronchoscopy 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Learning Objectives

After studying this chapter, you should be able to do the following:
  • Understand the indications for both respiratory and gastrointestinal (GI) endoscopy in critically ill patients.

  • Describe the contraindications and potential complications that are associated with endoscopy and endosurgery.

During the past three decades, the technology of endoscopy has led to the development of sophisticated optics and simple-to-use devices that are practical for use in critically ill patients. The upper airway, lung, and gastrointestinal (GI) passageways are accessible by endoscopy. In the intensive care unit, endoscopy is commonly performed for therapeutic reasons and diagnostic purposes. This chapter reviews the indications, contraindications, techniques, and complications of laryngoscopy, bronchoscopy, and GI endoscopy in critically ill patients.


The larynx can be examined directly with a laryngoscope or a bronchoscope. The traditional rigid laryngoscope is a lighted metal instrument that displaces the tongue and permits the observation of the glottis, including the epiglottis and vocal cords. It consists of a handle containing batteries and a detachable blade, either straight (Miller) or curved (McIntosh), which comes in a variety of sizes, and a bulb to illuminate the tip of the blade. The straight blade bypasses and lifts the epiglottis, while the curved blade tip fits into the vallecula (Fig. 10-1).
Figure 10-1

There are two basic types of laryngoscope blades, the MacIntosh (left) and the Miller (right) blades. The MacIntosh blade is curved and its use differs from that of the Miller blade. The MacIntosh blade tip is placed in the vallecula, and the handle of the laryngoscope is pulled forward at a 45° angle. This technique allows visualization of the epiglottis. With the Miller blade, which is straight, the tip is placed posterior to the epiglottis, and therefore, the epiglottis is pinned between the base of the tongue and the straight laryngoscope blade. The motion on the laryngoscope handle is the same as that used with the MacIntosh blade.

The upper airway can be inspected by both rigid and flexible laryngoscopy techniques.
Fiberoptic laryngoscopy has enhanced the ease and comfort of visualizing the oronasopharynx and larynx. The flexible fiberscope, inserted via the nose, is considered as the simplest method of visualizing the upper airway and an excellent way to monitor the epiglottic swelling to evaluate the need to intubate or perform a tracheostomy. Rigid laryngoscopy should be performed for upper airway inspection, particularly in patients who are suspected of having upper airway obstruction (Table 10-1). The major indication for rigid laryngoscopy is endotracheal intubation.
Table 10-1

Causes of Upper Airway Obstruction


Facial and neck injury

Laryngeal trauma and stenosis

Airway burn



Quinsey (acute tonsillitis)

Ludwig’s angina (mouth floor abscess)

Retropharyngeal abscess

Endotracheal tube trauma

Foreign body




Vocal cord paralysis

Postextubation stridor

Rigid laryngoscopy facilitates endotracheal intubation.

Endotracheal intubation with a flexible fiberoptic laryngoscope or bronchoscope requires training. First, the larynx must be localized by visualization. Second, the tip of the scope is advanced into the trachea. Excessive secretions and tissue edema can hamper visualization of the larynx. Although the maneuver may sound simple, it can be difficult.


Bronchoscopy is the endoscopic examination of the larynx and tracheobronchial tree. In the intensive care unit, it can be performed with a rigid or flexible bronchoscope. It is rarely performed using a rigid instrument, but is predominately performed using a flexible bronchoscope.

With the presence of bronchopulmonary disease in the intensive care unit and the use of mechanical ventilation with endotracheal tubes, bronchoscopy has become an important tool in the management of ICU patients. Fiberoptic bronchoscopy can be easily performed and is rarely associated with complications. With proper attention to patient sedation, it can be performed without significant patient discomfort. It has surpassed rigid bronchoscopy as the instrument of choice for evaluating the tracheobronchial tree. Further, as compared to rigid bronchoscopy, flexible bronchoscopy allows for more complete exploration of the airways. There is access to the upper lobes that rigid bronchoscopy does not allow. The procedure can easily be performed at the bedside, with minimal technical assistance.
Flexible fiberoptic bronchoscopy can be performed through an endotracheal tube while a patient is mechanically ventilated.
Flexible fiberoptic bronchoscopy allows the most complete inspection of the large airways.
Despite the ease of use of flexible fiberoptic bronchoscopy, the rigid bronchoscope still has advantages in certain clinical situations. Two such instances are massive hemoptysis and the removal of large foreign bodies. Although laser surgery can be performed using a flexible fiberoptic bronchoscope, the rigid bronchoscope has distinct advantages for this intervention. It also has an advantage in certain dilation procedures of the tracheobronchial tree involving airway stricture and tumors, and in particular, the placement of airway stents. Nonetheless, flexible fiberoptic bronchoscopy can also be used for these procedures.
Rigid bronchoscopy is most helpful in massive hemoptysis and the removal of a foreign body.

Diagnostic Indications

The diagnostic indications for bronchoscopy are numerous (Table 10-2). However, the most commonly encountered diagnostic indications in the ICU include the evaluation of hemoptysis, atelectasis, diffuse parenchymal disease, inspection of the airways following an inhalation injury or blunt trauma, assessment of the large airways following intubation, and for culturing airway secretions or washings, including bronchoalveolar lavage.
Table 10-2

Indications for Bronchoscopy

Diagnostic Indications

Therapeutic Indications

Acute inhalation injury

Airway stent placement

Assessment of intubation trauma

Bedside tracheostomy


Brachytherapy for central airway neoplasms

Blunt chest trauma

Closure of bronchopleural fistula

Chest radiograph consistent with neoplasia

Endotracheal intubation


Excessive secretions and atelectasis


Foreign bodies

Diaphragmatic paralysis


Diffuse parenchymal disease and/or bilateral hilar adenopathy

Laser resection for central airway neoplasms

Lung abscess

Preoperative assessment of resectability

Pulmonary alveolar proteinosis lavage



Laryngeal abnormalities


Localized wheeze


Lung abscess


Metastatic disease with unknown primary site


Pleural effusion of unknown etiology


Positive cytology and normal chest radiograph


Recurrent laryngeal nerve paralysis


Recurrent pneumonia


Symptoms after resection surgery


Unresolving infiltrate

Bronchoscopy is often used to evaluate hemoptysis and to culture airway secretions and lung washings in critically ill patients.

Bronchoscopy can be employed to identify the site of bleeding when a patient has hemoptysis. Hemoptysis can be caused by a variety of tracheobronchial lesions, cardiovascular and hematologic conditions as well as localized and diffuse parenchymal lung disorders. The clinician must decide when bronchoscopy is indicated in the patient with hemoptysis.

For example, the patient who has hemoptysis as a complication of pulmonary embolism does not likely require bronchoscopy. However, a respiratory failure patient with a localized lung infiltrate on chest X-ray and hemoptysis does require bronchoscopy to determine if there is an airway lesion. Diseases such as carcinoma can be responsible for the latter clinical presentation. In intubated patients, hemoptysis should always be evaluated to determine if tracheal damage has occurred at the time of or during intubation. The bronchoscopist should carefully inspect the airways and attempt to find the source of bleeding, when possible. The use of the fiberoptic bronchoscope allows the bronchoscopist to evaluate the areas to the segmental, and even subsegmental, bronchial level.
Hemoptysis following intubation should always be evaluated by bronchoscopy.

Many chest X-ray abnormalities can be evaluated by bronchoscopy. However, not all chest X-ray abnormalities require diagnostic bronchoscopy. Not even all intensive care patients with pneumonia require bronchoscopy. Some situations such as, when the patient is immunocompromised and/or special cultures are necessary can be clear indications for diagnostic bronchoscopy to obtain specimens that may be recovered only with the use of a bronchoscope. Certain chest X-ray abnormalities may indicate a need for diagnostic bronchoscopy (see Table 10-2). These abnormalities include atelectasis (of an entire lung, lobe, or segment), an enlarging or suspicious pulmonary parenchymal mass, cavitating pulmonary lesions, as well as diffuse parenchymal processes that do not have an established diagnosis.

Atelectasis often requires fiberoptic bronchoscopy to attempt to rule out endobronchial obstruction by carcinoma or a foreign body. Mucous plugging of the airway many times causes atelectasis. When atelectasis occurs in a critically ill patient who has had a normal chest X-ray on admission, mucous plugging is a common cause.1 In patients who are intubated, the position of the endotracheal tube can be responsible for the atelectasis. Endotracheal tubes can slip down a mainstem bronchus, generally the right main stem, and obstruct the right upper lobe creating right upper lobe atelectasis. Additionally, the left lung may not have been aerated during this process, and complete left lung atelectasis may occur.
Lung atelectasis often necessitates bronchoscopy to rule out an endobronchial lesion, a foreign body, or mucous plugging.
Bronchoscopy can be helpful in the diagnosis of both bacterial and nonbacterial pulmonary infections in critically ill patients.1 Lung secretions can be obtained by several different ways, including bronchial washings, bronchoalveolar lavage, protected-catheter brushings, and in select patients, transbronchial lung biopsy. Certain factors must be considered when choosing a certain bronchoscopy procedure in a critically ill patient with diffuse lung infiltrates. They include physician expertise, the patient’s condition, and the potential for diagnostic success of a particular procedure in a select patient.
Bronchoscopy can help diagnose a variety of pulmonary infections.
The two major bronchoscopic procedures used in the diagnosis of diffuse lung disease are bronchoalveolar lavage and transbronchial lung biopsy. Often, the protected-brush catheter (Fig. 10-2) is employed with bronchoalveolar lavage in the diagnosis of lung infection. Quantitative culturing using these techniques has improved the diagnosis of infection of the lung. Cultures obtained from the protected-brush catheter that provides 103 or more colony-forming units (CFU) per milliliter indicate active infection.2 This brush catheter technique is equivalent to needle aspiration biopsying of the lung in the identification of the etiology of bacterial pneumonia.2 The protected-brush catheter technique is limited by the fact that it only samples a very small area. Cultures obtained from this area may not alter the antibiotic therapy that is being used at that time because of the use of broad-spectrum antibiotics that treat most of the common bacterial infections. Bronchoalveolar lavage, performed through the flexible bronchoscope when wedged peripherally in an airway of the lung, allows for the recovery of both cellular and noncellular components of the lower respiratory tract. Bronchoalveolar lavage has been used to diagnose certain interstitial lung diseases, malignancies, and infections. Quantitative cultures obtained from the bronchoalveolar lavage that provides 104 or more CFU per milliliter indicate active infection.3 Unfortunately, bronchoalveolar lavage is not the preferred method to make a definitive diagnosis for most interstitial lung diseases, and tissue biopsy is the diagnostic procedure of choice for that set of diseases. Bronchoalveolar lavage has been very successful in the diagnosis of Pneumocystis carinii pneumonia and other causes of diffuse pulmonary infiltrates in immunocompromised hosts.4 Its greatest efficacy is in the diagnosis of opportunistic infections, including Pneumocystis carinii, cytomegalovirus, a variety of fungi, and mycobacterium.5 The detection of hemosiderin-laden macrophages can be helpful in the diagnosis of pulmonary hemorrhage. Occasionally, it may be helpful in the diagnosis of pulmonary malignancy, particularly in those patients who have lymphangitic metastasis.6
Figure 10-2

The plugged telescoping catheter brush is used to obtain selective samples from the lower airways and to keep the brush sterile until it is pushed out of the catheter at the time of culturing.

Protected-brush catheter and bronchoalveolar lavage techniques are useful in the diagnosis of pneumonia.
Bronchoalveolar lavage is the diagnostic procedural choice for Pneumocystis carinii pneumonia.
Critically ill patients who require fiberoptic bronchoscopy for airway inspection include those who have sustained serious inhalation injuries or blunt chest trauma and those patients suspected of having intubation damage. For patients who have sustained an inhalation injury, the presence of serious mucosal injury can be identified during fiberoptic bronchoscopy. A decision for prophylactic intubation can be better addressed with this information available. Fiberoptic bronchoscopy allows the trauma surgeon to evaluate the airways following blunt chest trauma to determine if the patient has sustained a fractured airway. This is suspected if atelectasis, pneumomediastinum, or pneumothorax is determined during the evaluation. Fiberoptic bronchoscopy can be used to determine whether laryngeal or tracheal complications have occurred during an intubation. If serious injury has occurred, then a tracheostomy should be considered. Fiberoptic bronchoscopy can be performed in patients who are intubated and have an endotracheal tube in place that is 7.5 mm or larger in internal diameter. Therefore, it is important for most of our adult patients to use an endotracheal tube at least 8 mm in internal diameter during intubation in case bronchoscopy is needed. A complete airway inspection can be done with a fiberoptic bronchoscope. With the endoscope inserted through the endotracheal tube, the balloon can be deflated and the tube withdrawn over the bronchoscope to look for subglottic damage. The tube can be carefully withdrawn up through the vocal cords over the fiberoptic bronchoscope for glottic and supraglottic assessment. The presence of serious mucosal ulceration, necrosis, or edema indicates the need for a tracheostomy. Tracheostomy lessens the likelihood of the consequences of tracheomalacia, tracheal stenosis, and laryngeal stenosis.
Following blunt chest trauma, bronchoscopic inspection can determine airway fracture and laryngeal injury.

Therapeutic Indications

Many bronchoscopies in the intensive care unit are performed for airway secretion management. It is common to perform bronchoscopy for both diagnostic and therapeutic reasons in patients who are intubated and are critically ill. The therapeutic uses of fiberoptic bronchoscopy (see Table 10-2) are as important as its diagnostic indications.
Airway secretion management can be facilitated by bronchoscopy in intubated patients.
When aggressive pulmonary toilet, including physical therapy, incentive spirometry, and sustained maximum inspiration with cough, fails to clear the airways of excessive secretions or reexpand significant lung atelectasis, then fiberoptic bronchoscopy can be considered. Retention of secretions and mucous plugging of the airways are common clinical complications in those patients with an altered level of consciousness and impaired cough. Poor pulmonary function often results from weakness, recurrent aspiration, ventilator dependence, or pain following thoracoabdominal surgery. Following thoracic surgery, blood clots often accumulate in the lung airways and can induce atelectasis. Patients with airway mucosal injury are more likely to have serious secretion and mucous plugging problems. In these situations, fiberoptic bronchoscopy may enhance pulmonary toilet and can be lifesaving.
Bronchoscopy can remove retained secretions, mucous plugs, blood clots, and foreign bodies.

For airway secretion management, it is better to use a flexible bronchoscope with large-channel suctioning capabilities. The airway secretions can be thick and tenacious making it difficult to remove. In an intubated patient, the fiberoptic bronchoscope can be introduced to enhance secretion removal. Occasionally, the installation of N-acetylcysteine or pulmozyme through the bronchoscope may be necessary to help liquify thick, tenacious inspissated mucus from the airways. The use of acetylcysteine may trigger bronchospasm, but patients who experience this complication generally respond well to a bronchodilator nebulizer treatment.7 Dornase alfa inhalation may cause less bronchospasm.

Fiberoptic bronchoscopy can be used to retrieve a foreign body lodged in an airway. Various techniques can be used either to grasp or net the object and to pull it out of the airway with the bronchoscope when it is withdrawn. Fiberoptic bronchoscopy can be used for endotracheal intubation. The bronchoscope acts as an obturator for endotracheal intubation in patients for whom intubation is difficult. In select patients with neck or head trauma or certain disease states such as ankylosing spondylitis or in those who have had laryngeal trauma or have vocal cord dysfunction, this technique can be particularly helpful. With the endotracheal tube over the bronchoscope, the bronchoscope is advanced either through the nose or orally to the vocal cords under direct visualization. After placing the endoscope through the cords, the endotracheal tube can be slipped over the bronchoscope into the airway. This technique can be helpful in individuals with massive facial injuries.
Bronchoscopy can facilitate endo- or nasotracheal intubation.
Fiberoptic bronchoscopy can assist in the treatment of massive hemoptysis. Bronchial tamponade can be accomplished with a transbronchoscopic endobronchial balloon occlusion technique after the bronchoscope has been placed in an airway that is hemorrhaging. Bronchial tamponade can be used for either bronchial or pulmonary hemorrhage, as well as for refractory pneumothorax secondary to persistent air leaks following thoracotomy.
Bronchoscopy helps locate the area of hemoptysis, and it can assist in tamponade.

Sometimes a bronchopleural fistula can develop between a bronchial tree and a pleural space. After chest tube placement, bronchoscopy can identify which airway is part of this phenomenon and then be used to distally occlude the bronchopleural fistula. A variety of materials have been injected through the bronchoscope to seal the fistula.8

Central airway-obstructing lesions involving the larynx, trachea, or a major bronchus can be treated using a variety of bronchoscopic techniques including photoresection using laser technology or airway stenting.9 Airway stenting can be used in malignant or benign disease with severe airway narrowing from intrinsic or extrinsic processes. The trachea and main stem bronchi can be stented, but the technique is not suitable for lobar and distal bronchial stenosis. Many types of tracheobronchial stents are available, such as expandable metal wire (Fig. 10-3), molded silicon stents, or a combination of these. After the area of stenosis has been identified and balloon dilatation is used, then a stent is placed in that area and distended, expanding the airway lumen and relieving the critical stenosis. Complications can occur with balloon dilation and stenting, such as airway rupture, stent migration, increased mucosal secretions, and a granulomatous mucosal reaction.
Figure 10-3

(ac) Wire stent dilation of airway. (a) A guidewire is placed into the distal end of the delivery catheter. Under direct visualization with the bronchoscope, the catheter is threaded over the guidewire into the stricture. (b) The wire stent is expanded under fluoroscopic guidance to dilate the airway stricture. (c) The catheter and bronchoscope are withdrawn after the airway is dilated open.

Airway stenting can open stenotic segments of large airways.


When performed by a trained physician, flexible fiberoptic bronchoscopy has a very low morbidity and mortality incidence. Mortality does not exceed 0.1%, and overall complications should not exceed 10%.10 The complications may be related to the use of procedure premedications including sedatives or topical anesthetics, and vagal-mediated reactions or complications from the bronchoscopy itself or its related procedures. Absolute contraindications to performing bronchoscopy include an unstable cardiovascular status including life-threatening cardiac arrhythmias, severe hypoxemia that is likely to worsen during the procedure, and an inexperienced bronchoscopist and bronchoscopy team. Coagulation problems are not an absolute contraindication to bronchoscopy when a tracheobronchial inspection is needed or only bronchoalveolar lavage is to be performed. However, more invasive procedures including biopsy and needle aspiration techniques should not be performed until the coagulopathy has been corrected.
Serious complications are rare with bronchoscopy.

Premedication with sedatives may lead to respiratory depression, hypoventilation, hypotension, and syncope. Topical anesthetics can cause laryngospasm, bronchospasm, and if given excessively, seizures. The bronchoscopy itself or its associated procedures such as brushing, biopsy, and bronchoalveolar lavage may also induce laryngospasm, bronchospasm, hypoxemia, and cardiac arrhythmias or cause fever, pneumothorax, and hemorrhage. Patients on mechanical ventilation who are receiving high positive end-expiratory pressure (PEEP) are at increased risk for barotrauma. Although fever is a relatively common complication, it is generally transient and not associated with sustained bacteremia. Perforation of the airway is exceedingly uncommon, but pneumothorax has been reported.

Procedure for Patient Preparation

Although each institution is different, certain guidelines should and do exist. As in all medical procedures, patient preparation prior to the procedure can help to enhance the success and minimize any complications and adverse outcomes (Table 10-3). Understanding of the patient’s history and current condition is paramount to successful bronchoscopy.
Table 10-3

Patient preparation for Fiberoptic bronchoscopy (fob)

1. Obtain a list of all current medications and verify if the patient has held:

(a) Aspirin for 5 days.

(b) Plavix (clopidogrel) for 5 days.

(c) Aggrenox for 5 days.

(d) Ticlid (ticlopidine) for 5 days.

(e) Coumadin (warfarin) for a minimum of 3 days. Prior to procedure an INR of <1.3 must be documented.

(f) Heparin stopped for 24 h. Prior to procedure a normal PTT must be documented.

(g) Low molecular weight heparin stopped for 24 h.

2. Document a normal PT/INR, PTT, and platelet count within 7 days of the procedure. If the patient is on Coumadin (warfarin) or heparin, a normal coagulation profile (INR < 1.6) must be documented off these anticoagulants prior to the procedure.

3. Document that the patient has held NPO at least 4 and preferably 6 h prior to the procedure. Hypoglycemic medications are generally held prior to the procedure, as the patient is NPO.

4. Verify FOB indication and procedural plan.

5. Document the patient’s known allergy history.

6. Verify that the patient or responsible party has been explained about the procedure and informed consent has been appropriately filled out, signed, and witnessed.



Generally, before bronchoscopy, an anxiolytic, antisialagogue, and topical anesthetic are administered. Either the oral or nasal route can be used for flexible fiberoptic bronchoscopy. In the intensive care unit, the procedure is generally done through the endotracheal tube while the patient is on the mechanical ventilator. This procedure is done with the use of a swivel adapter with a rubber diaphragm through which the bronchoscope can be inserted safely without disconnecting the patient from the ventilator. The procedure should be done when the patient is as stable as possible. For critically ill patients, careful monitoring, including the use of oximetry and electrocardiography are necessary.

When bronchoscopy is performed through an endotracheal tube, the tube should have a lumen of at least 8 mm in internal diameter to help ensure that the patient receives an adequate tidal volume and that excessive airway resistance and pressure do not develop. While the procedure is being done, the patient generally receives 100% oxygen to optimize the chances of keeping the arterial oxygen saturation greater than 90%. The procedure should be done as quickly as possible but without compromising thoroughness. It is best to perform the procedure in patients who have fasted or at least after the stomach has been emptied of its contents by gastric tube suctioning. In mechanically ventilated patients, after the procedure has been completed, it is best to return the ventilator settings to the preprocedure settings and to obtain a chest X-ray to look for a postprocedure pneumothorax. Although postbronchoscopy fever develops in approximately 15% of patients, it usually lasts less than 24 h and does not require antibiotic therapy. However, if the fever does last more than 24 h, then postprocedure pneumonia should be considered.11

Gastrointestinal Endoscopy

At the present time, GI endoscopes are capable of visualizing nearly 100% of the upper GI tract, including the esophagus, stomach, and duodenum. Enteroscopes allow examination of the proximal small intestine, and the colon and terminal ileum are also accessible to examination with video-colonoscopes. The portability of these scopes allows these procedures to be performed on critically ill patients in the intensive care unit. The scopes are thin and flexible and are usually well tolerated with conscious sedation, even in critically ill patients. However, as with all interventions, the potential benefits for diagnosis and therapy must be weighed against the risks for each individual patient. This section reviews the indications, contraindications, and complications of GI endoscopy in critically ill patients.
GI endoscopy is done for both diagnostic and therapeutic reasons.


The indications for GI endoscopy in the intensive care unit are found in Table 10-4. If the clinician believes that the patient is too critically ill to be taken to the GI endoscopy suite, then gastroscopy and colonoscopy may be done with careful monitoring in the intensive care unit. The most frequent indication for emergency gastroscopy is upper GI bleeding. This can occur in two settings. The first is the patient who is admitted to the hospital for the primary diagnosis of GI bleeding. Initial assessment determines whether the patient is admitted for treatment to a general medical floor or to the intensive care unit. This decision is based on several risk factors in the patient with GI bleeding that are recognized to be associated with increased morbidity and mortality such as increased age, shock, comorbid illness, and the active passage of red blood per mouth or per rectum.
Table 10-4

Indications for Gastrointestinal (GI) Endoscopy

Upper GI endoscopy

GI bleeding

Caustic ingestion

Foreign body ingestion

Feeding tube placement

Endoscopic retrograde cholangiopancreatography (ERCP)

Severe gallstone pancreatitis

Severe cholangitis

Lower GI endoscopy

GI bleeding

Acute colonic ileus

GI endoscopy is commonly done for acute upper GI bleeding in critically ill patients.

The second situation in which upper GI bleeding is evaluated and treated in the intensive care unit is when bleeding develops in a patient already being treated in the intensive care unit for other medical or surgical reasons. The in-hospital mortality rate for patients in the latter situation is much greater than for patients admitted to the hospital for GI bleeding. The overall mortality for patients hospitalized for upper GI bleeding is 10%. However, newer evidence shows mortality rates for patients with upper GI bleeding complicating hospitalization for treatment of other illnesses. The in-hospital mortality in these two studies ranged from 42 to 77%. Both the studies found that the cause of death was usually from multisystem organ failure or sepsis related to the initial illness and not from blood loss. Endoscopic intervention frequently achieved hemostasis;however, rebleeding occurred in 30% of patients and repeat endoscopy was often needed.12

In some studies, peptic ulcer accounted for greater than 50% and stress gastritis or erosions caused approximately 20% of the bleeding episodes that arose during the treatment of other medical or surgical illnesses. One study found that 45% of patients had been treated during the hospitalization with aspirin or nonsteroidal antiinflammatory drugs prior to the onset of bleeding and 42% had been receiving corticosteroids.13 Aspirin and nonsteroidal antiinflammatory drugs have been proven to be the causative factors in the formation of ulcers and erosions in the stomach and duodenum and are associated with bleeding. The role of corticosteroids is more controversial. One study found that 44% of the patients were treated with mechanical ventilation and over 60% had a coagulopathy with an elevated prothrombin time or thrombocytopenia before the onset of bleeding.13 The latter finding highlights another observation that has been found in several studies that the risk of bleeding from stress ulceration or gastritis has been shown to be most closely associated with two risk factors; mechanical ventilation and coagulopathy. The presence of renal insufficiency in patients with either of these risk factors further increases the risk of bleeding.14 Patients with burns and acute central nervous system diseases are also at increased risk of bleeding from stress ulceration. Four percent of intensive care patients who require mechanical ventilation or have coagulopathy develop clinically important bleeding resulting in a decrease in hemoglobin greater than 2 g and associated with a fall in blood pressure. In contrast, only 0.1% of patients without these risk factors develop significant bleeding. This underscores the need for pharmacologic prophylaxis of stress gastritis in this group of patients. A proton pump inhibitor is most commonly used. General medical support to maintain adequate blood pressure and perfusion and the use of enteral nutritional support also probably reduces the risk of significant bleeding from stress ulceration, but no controlled trials of this nature exist.

In the same way that upper GI bleeding can be the primary reason for admission of the patient to the intensive care unit or a complication of a patient hospitalized for other reasons, the same is true for lower GI bleeding. The two most common causes of significant lower GI bleeding requiring admission to the hospital and treatment in the intensive care unit are colonic diverticulosis and colon angiodysplasia. However, the etiology of lower GI bleeding that complicates the treatment of other medical or surgical illness is unlikely to be either diverticulosis or angiodysplasia. In a study from Taiwan, it was found that, in this setting, lower GI bleeding was more commonly secondary to ischemic colitis, rectal ulcer, or pseudomembranous colitis.15 These complications are often seen in critically ill patients being treated in an intensive care unit since many were being treated for respiratory failure, sepsis, stroke, or circulatory compromise secondary to cardiac disease, and even some required therapy with antibiotics or vasopressor agents before the bleeding episode. In many instances, the underlying illness or its treatment predisposes to these complications. The in-hospital mortality rate of patients with lower GI bleeding as a complication of other medical illnesses is high, and in this study, was found to be greater than 50%.15 This is similar to the findings in upper GI bleeding complicating other illnesses where the cause of death is rarely secondary to hemorrhage. Other factors such as sepsis, respiratory failure, and multiorgan failure are the common causes of death.

Colonoscopy is difficult to perform during an episode of acute lower GI bleeding. Unlike gastroscopy, which can be successful in defining the site of blood loss during an episode of upper GI bleeding, colonoscopy is frequently a suboptimal exam and more often fails to determine the cause of bleeding. It can be difficult to adequately purge the colon in preparation for colonoscopy and residual blood and stool interferes with the passage of the scope and compromises the examination of the mucosa. The poor visualization during the exam can also increase the risk of complications during the procedure. In one study, colonoscopy was aborted in 20% of the cases because of an inadequate prep and too much blood in the lumen of the colon. The ability to perform a complete exam with the passage of the scope to the cecum was successful in only 60% of patients.15 This is in contrast to greater than 95% of elective colonoscopies. However, the cause of bleeding was frequently found in the left colon enabling a diagnosis to be established in two thirds of the cases. This study concluded that colonoscopy should be considered as a modality, in addition to angiography, to be used in the evaluation of lower GI bleeding.
Colonoscopy is done to locate the site of lower GI bleeding and is also used to decompress a markedly dilated colon.

Another indication for colonoscopy in the intensive care setting is for decompression of a dilated colon secondary to an ileus. Ileus is a common complication seen in patients critically ill with intraabdominal processes and in patients with systemic disease. It can be secondary to electrolyte abnormalities, such as hypokalemia, hypercalcemia and sodium disturbances. It can also be caused by medications such as narcotics or drugs with anticholinergic effects. Ileus may also develop in patients with neurologic disorders such as stroke or intracranial bleed or infection, and also occurs in patients with sepsis. Although ileus usually affects both the small intestine and colon, it can sometimes affect only the colon. This is also referred to as colonic pseudo-obstruction because the colon is dilated without concomitant distention of the small intestine, a radiographic finding that is seen with distal colonic obstruction. Colonic ileus can lead to significant dilation of the colon that can result in ischemia and perforation, most commonly in the cecum. Colonic ileus is diagnosed when the transverse colon is greater than 5 cm or the cecum is greater than 8 cm in diameter on an abdominal X-ray and distal colonic obstruction must be excluded. If the cecum dilates to greater than 10–12 cm, the risk of perforation significantly increases. Predisposing factors, such as electrolyte abnormalities or medications or infection, should be addressed and corrected if possible and decompression with naso-gastric and rectal tubes should be attempted. If not successful, a trial of intravenous neostigmine, a cholinesterase inhibitor, can be attempted after colonic obstruction is excluded. If successful, a response with the passage of gas and fecal residue and improvement of distention is usually seen within minutes.16 If colonic dilation persists, then colonoscopy for decompression can be attempted at the bedside. Although colonoscopy may successfully decompress the colon, until the etiology of the colonic ileus is corrected, there remains a likelihood of recurrent colonic distention. Therefore, a drainage tube can be passed with the colonoscope and left in place in the proximal colon for continued decompression to try to reduce recurrent dilation. Because no prep is possible before the colonoscopy in this clinical situation, the procedure is difficult and carries an increased risk of complications.

Similar to the varied indications for colonoscopy, in addition to the diagnosis and treatment of GI bleeding, there are other indications for gastroscopy in the intensive care patient. These include the removal of foreign bodies, evaluation of caustic ingestion, and endoscopic placement of percutaneous gastrostomy feeding tubes.

The endoscopic placement of a percutaneous gastrostomy feeding tube affords a route for enteral nutrition that reduces these risks and provides access to the GI tract for feeding and for medications. After initiating enteral feeds through the gastrostomy tube, residual gastric contents must be checked to assess gastric emptying and minimize the risk of aspiration.
A gastrostomy tube can be placed percutaneously by an endoscopic technique at the bedside.
ERCP occasionally is necessary in an ICU patient. ERCP should be considered in a patient who has severe gallstone pancreatitis or in a patient with cholangitis unresponsive to medical therapy. ERCP combined with sphincterotomy and gallstone extraction reduces the complications that can occur in patients who have gallstone pancreatitis and cholangitis. For the patient with cholangitis, biliary stents can be placed if common duct stones cannot be removed during the procedure.
Endoscopic retrograde cholangiopancreatography (ERCP) is indicated for the patient with cholangitis that is unresponsive to medical therapy.


Bleeding and perforation are the major risks of GI endoscopy (Table 10-5). Bleeding can occur as a complication of endoscopic intervention with thermal coagulation. Perforation of the bowel wall may result from direct pressure of the endoscope against the wall or by the inadvertent puncture of the wall by an endoscopic instrument. Perforation may also complicate the use of thermal coagulation used to stop bleeding. Aspiration of gastric contents during gastroscopy is always a concern and this risk is increased during active upper GI bleeding. Elective endotracheal intubation may be done prior to endoscopy to protect the airway when the risk of aspiration is felt to be high. There may also be complications related to sedation such as respiratory depression and hypotension. The best way to avoid serious procedural complications is to avoid the patient who has certain contraindications to the procedure (Table 10-6).
Table 10-5

Complications of GI Endoscopy


Perforation of the GI tract lumen by endoscope, catheters, or guidewires


Reaction to sedative medication

Table 10-6

Contraindications to GI Endoscopy

Perforated viscus suspected/impending

Hemodynamic and respiratory gas exchange instability

Severe diverticulitis

Severe inflammatory bowel disease

Severe coagulopathy

Uncooperative patient

Unprotected airway in a confused or stuporous patient with acute upper GI bleeding

Major risks of colonoscopy are bleeding and perforation.
In addition to these general complications of endoscopic procedures, individual procedures are associated with specific risks. ERCP and sphincterotomy carries an increased risk of perforation or bleeding during sphincterotomy and also may cause pancreatitis. ERCP that is done to decompress an obstructed biliary tree, if unsuccessful, can worsen cholangitis and cause bacteremia and septicemia. Percutaneous endoscopic gastrostomy tube placement can be complicated by inadequate formation of the feeding tube tract between the abdominal wall and the stomach. This can result in peritonitis. This risk is the greatest in patients who are severely malnourished with low serum albumin and in patients receiving corticosteroids. Wound infection at the gastrostomy tube site and leakage of gastric contents through the site are other complications that are also increased in critically ill patients, especially those who are malnourished. Unique complications can also occur during the treatment of esophageal varices (Table 10-7). Sclerosis of esophageal varices is associated with complications including chest pain, fever, esophageal ulceration, bleeding, stricture, and perforation. Pleuro-pulmonary complications may develop and mediastinitis can rarely occur. Variceal banding can cause superficial esophageal ulceration, but these ulcers usually do not bleed, nor do they result in the other complications that are seen with sclerosis.
Table 10-7

Complications of Endoscopic Coagulation and Injection Sclerotherapy

Esophageal complications


Stricture formation

Perforation (early or delayed)


Pulmonary complications

Pleural effusions and pleuritis

Pulmonary infiltrates



Adult respiratory distress syndrome

Septic complications



“Spontaneous” bacterial peritonitis


During episodes of bleeding, upper and lower GI endoscopy should be done following fluid resuscitation. Naso-gastric lavage with large-bore tube irrigation should be performed to empty the stomach of blood before gastroscopy. The endoscopy team should consist of an experienced endoscopist, and an assistant skilled in monitoring the patients who are undergoing endoscopy.

A variety of endoscopic techniques can be used to control bleeding. These include laser photocoagulation, thermal electrocoagulation, injection therapy, hemoclipping, and ligation. Laser equipment is expensive and not easily portable and is rarely used at the bedside. In contrast, the other techniques are readily available and can be done in the intensive care unit. Directed injection to the bleeding site using a catheter with a needle tip that is passed through the endoscope offers the ability to give injection therapy and control bleeding. Sclerosing agents can be injected into varices to control bleeding. For other bleeding lesions, such as ulcers or Mallory Weiss tears, epinephrine can be injected as a temporizing technique to constrict the bleeding vessel. This reduces or stops bleeding and facilitates more definitive endoscopic treatment such as thermal coagulation. Multipolar probes or heater probes can be passed through the endoscope to stanch bleeding by coagulating the bleeding vessel. Endoscopic clips can also be applied to compress the bleeding vessel to control bleeding. Because of the complication profile, variceal sclerosis has been replaced by band ligation as the favored approach of most endoscopists in the treatment of bleeding esophageal varices. However, sclerosis is still utilized for bleeding gastric varices and by some endoscopists as an alternative to banding for esophageal varices. It is also used if banding ligation fails to stop variceal bleeding.


Endoscopy has substantially broadened our diagnostic and therapeutic powers in the intensive care unit. Laryngoscopy, bronchoscopy, and GI endoscopy are rarely contraindicated and often simplify the care of patients who are seriously ill.

Review Questions

  1. 1.

    All the following statements are true concerning bronchoscopy, except:

  2. A.

    Flexible bronchoscopy is commonly done in the ICU

  3. B.

    Rigid bronchoscopy does not have any advantages over flexible bronchoscopy

  4. C.

    Bronchoscopy is safely done often with minimal additional sedation in mechanically ventilated patients

  5. D.

    The most common use of bronchoscopy in the ICU is for secretion management

  6. 2.

    The fiberoptic bronchoscopy is helpful in the ICU for all of the following conditions, except:

  7. A.

    Location of bleeding

  8. B.

    Diagnosis of hospital-acquired pneumonia

  9. C.

    Removal of an airway mucous plug

  10. D.

    Diagnosis of inflammatory lung diseases

  11. 3.

    All the following statements are true concerning GI endoscopy in the ICU, except:

  12. A.

    Nearly the entire GI tract can be evaluated

  13. B.

    All GI problems should be considered for endoscopy evaluation

  14. C.

    The most common cause for using endoscopy is bleeding

  15. D.

    Bleeding and bowel perforation are the major risks of GI endoscopy



  1. 1.

    The answer is B. Massive hemoptysis, the removal of a foreign body, and the placement of an airway stent are often best approached with a rigid bronchoscope.

  2. 2.

    The answer is D. The bronchoscopic approach to the diagnosis of diffuse inflammatory noninfectious lung diseases is often unrewarding and is better approached by an open-lung biopsy, either by thoroscopy or thoracotomy.

  3. 3.

    The answer is B. Clinically insignificant GI problems should not be evaluated endoscopically. All the patients who require endoscopy should be as stable as possible.



  1. 1.
    Prakash U. Bronchoscopy in the critically ill patient. Semin Respir Med. 1997;18:583-591.CrossRefGoogle Scholar
  2. 2.
    Fagon JY, Chastre J, Hance AJ. Detection of noscomial lung infection in ventilated patients: use of a protected specimen brush and quatitative culture techniques in 147 patients. Am Rev Respir Dis. 1998;138:110-116.CrossRefGoogle Scholar
  3. 3.
    Gerbeaux P, Ledoray V, Boussuges A. Diagnosis of nosocomial pneumonia in mechanically ventilated patients: repeatability of the bronchoalveolar lavage. Am J Respir Crit Care Med. 1998;157:76-80.PubMedGoogle Scholar
  4. 4.
    Golden JA, Hollander H, Stulbarg MS, Gamsu G. Bronchoalveolar lavage as the exclusive diagnostic modality for pneumocystis carinii pneumonia. A prospective study among patients with acquired immunodeficiency syndrome. Chest. 1986;90(1):18-22.PubMedCrossRefGoogle Scholar
  5. 5.
    Raoof S, Rosen J, Khan F. Role of bronchoscopy in AIDS. Clin Chest Med. 1999;20(1):63-76.PubMedCrossRefGoogle Scholar
  6. 6.
    Rennard S. Bronchoalveolar lavage in the diagnosis of cancer. Lung. 1990;168(1):1035-1040.PubMedCrossRefGoogle Scholar
  7. 7.
    Bernstein I, Ausdenmoore R. Iatrogenic bronchospasm occurring during clinical trials of a new mucolytic agent, acetylcysteine. Dis Chest. 1964;46:469-473.PubMedCrossRefGoogle Scholar
  8. 8.
    Chang C, Hsu H, Kuo S, Lee Y. Bronchoscopic gluing for post-lung-transplant bronchopleural fistula. Eur J Cardiothorac Surg. 2007;31(2):328-330.PubMedCrossRefGoogle Scholar
  9. 9.
    Cavalieri S, Foccoli P, Feijo S. Nd:YAG laser therapy in lung cancer. An 11-year experience with 2,253 applications in 1,585 patients. J Bronchol. 1994;1:105-111.CrossRefGoogle Scholar
  10. 10.
    Dellinger RP, Band V. Fiberoptic bronchoscopy in the intensive care unit. Crit Care Clin. 1992;8:755-772.PubMedGoogle Scholar
  11. 11.
    Shennib H, Baslaim G. Bronchoscopy in the intensive care unit. Chest Clin North Am. 1996;6:349-361.Google Scholar
  12. 12.
    Lee Y, Wang H, Ming-Shang W. Urgent bedside endoscopy for clinically relevant gastrointestinal hemorrhage after admission to the intensive care unit. Intensive Care Med. 2003;29:1723-1728.PubMedCrossRefGoogle Scholar
  13. 13.
    Lewis J, Shin E, Metz D. Characterization of gastrointestinal bleeding in severely ill hospitalized patients. Crit Care Med. 2000;28:46-50.PubMedCrossRefGoogle Scholar
  14. 14.
    Klebl F, Scholmerich J. Therapy insight: prophylaxis of stress induced gastrointestinal bleeding in critically ill patients. Nat Clin Pract Gastro Hepat. 2007;4:562-570.CrossRefGoogle Scholar
  15. 15.
    Chun-Che L, Yi-Chia L, Huei L. Bedside colonoscopy for critically ill patients with acute lower gastrointestinal bleeding. Intensive Care Med. 2005;31:743-746.CrossRefGoogle Scholar
  16. 16.
    Ponec R, Saunders MD, Kimmey MB. Neostigmine for the treatment of acute colonic pseudo-obstruction. NEJM. 1999;341: 137-141.PubMedCrossRefGoogle Scholar

Additional Reading

  1. Anzueto A, Levine SM, Jenkinson SG. The technique of fiberoptic bronchoscopy. Diagnostic and therapeutic use in intubated, ventilated patients. J Crit Illness. 1992;7:1657-1664.Google Scholar
  2. Dellinger P, Bandi V. Fiberoptic bronchoscopy in the intensive care unit. Crit Care Clin. 1992;8:755-772.PubMedGoogle Scholar
  3. Hirschchowitz BI. Development and application of endoscopy. Gastroenterology. 1993;104:337.Google Scholar
  4. NIH Consensus Conference. Therapeutic endoscopy and bleeding ulcers. JAMA. 1989;262:1369.Google Scholar
  5. Shennib H, Baslaim G. Bronchoscopy in the intensive care unit. Chest Surg Clin N Am. 1996;6:349-361.PubMedGoogle Scholar
  6. Steer ML, Silen W. Diagnostic procedures in gastrointestinal hemorrhage. N Engl J Med. 1983;309:646-650.PubMedCrossRefGoogle Scholar
  7. Van Dam J, Brugge WR. Endoscopy of the upper gastrointestinal tract. N Engl J Med. 1999;341:1738-1748.PubMedCrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC 2010

Authors and Affiliations

  • Harvey M. Licht
    • 1
  • Fredric Jaffe
    • 2
  • Gilbert E. D’Alonzo
    • 3
  1. 1.Department of MedicineTemple University School of MedicinePhiladelphiaUSA
  2. 2.Department of Medicine, Division of Pulmonary and Critical CareTemple University HospitalPhiladelphiaUSA
  3. 3.Pulmonary and Critical Care MedicineTemple University School of MedicinePhiladelphiaUSA

Personalised recommendations