**2. Sedation-related complications of colonoscopy**

Sedation-related complications of colonoscopy are usually cardiovascular, pulmonary and oc‐ casionally neurological. The risk of these events occurring is associated with advancing age, higher American Society of Anesthesiologists Physical Status Classification System scores (ASA score—with category 6 not being applicable), and the patient's co-morbidities [6-8].

ASA Physical Status Classification System (I-VI)

ASA-I A normal healthy patient.

gories of complications associated with colonoscopy are widely recognized. They are seda‐ tion-related complications, hemorrhage associated with colonic polypectomy and colonoscopy-related colonic perforation. Sedation-related complications are usually cardio‐ vascular and/or pulmonary and include oxygen desaturation, respiratory arrest, alterations in heart rate (bradycardia and tachycardia), cardiac arrhythmias, myocardial infarction, stroke, seizures (at times attributed to the method of preparation) and shock. Hemorrhage is most often associated with snare electrocautery polypectomy but may also occur during the performance of diagnostic colonoscopy with or without biopsies. Two general subcategories of hemorrhage exist: hemorrhage immediately following the performance of polypectomy or delayed hemorrhage occurring up to several weeks after the therapeutic procedure. Co‐ lonic perforation resulting from colonoscopy may occur due to mechanical forces exerted against the colonic wall (colonoscope tip or shaft, biopsy forceps, dilatation of a stricture), barotrauma as a result of intraluminal air or carbon dioxide insufflation, or as a result of a therapeutic procedure such as polypectomy, foreign body extraction, or stent placement to name a few. A thorough understanding of these complications, their incidence and treat‐ ment, is part of the training of all those learning to perform colonoscopy and forms the basis for the physician obtaining informed consent (an explanation of the risks and benefits of the procedure) from the patient. This chapter will systematically review our current under‐ standing of these complication categories and the methods of minimizing the likelihood of developing these complications. The latest treatments of specific complications will be re‐ viewed with the intent of aiding the physician endoscopist's understanding of the principles

*"Primum non nocere"* is the Latin phrase that means "First, do no harm". Non-maleficence, which is derived from this maxim, is one of the principal precepts of medical ethics taught to all medical students in medical school and is a fundamental principle for the provision of medical services world-wide. Another way to state it is that "given an existing problem, it may be better not to do something, or even to do nothing, than to risk causing more harm than good." It reminds the physician and other health care providers that they must consid‐ er the possible harm that any intervention might do. It is invoked when debating the use of an intervention that carries an obvious risk of harm but a less certain chance of benefit. This ancient principle should be kept in mind when contemplating colonoscopy and the possible

The doctrine of informed consent (and its antithesis, informed refusal) for colonoscopy in‐ volves an assessment of the competence of the patient by the physician, disclosure of, in an understandable way, the information necessary to allow the patient to make an informed decision (risks and benefits considered) regarding the role of colonoscopy in his care, and the documentation of these proceedings in the medical record [3]. It is an intrinsic part of the doctor-patient relationship and an ethical obligation on the part of the physician in the clini‐ cal practice of medicine. In the United States, the doctrine of medical informed consent is most famously traced back to a 1914 New York court decision centered about the observa‐

of risk management as regards to performing colonoscopy.

198 Colonoscopy and Colorectal Cancer Screening - Future Directions

complications of the procedure.

**1.1. Informed consent for colonoscopy**

ASA-II A patient with mild systemic disease.

ASA-III A patient with severe systemic disease.

ASA-IV A patient with severe systemic disease that is a constant threat to life.

ASA-V A moribund patient who is not expected to survive without the operation.

ASA-VI A declared brain-dead patient whose organs are being removed for donor purposes.

ASA-E Emergency operation of any variety (used to modify one of the above classifications, i.e., ASA III-E)

In general, patients' inpatient status, trainee participation and the routine use of supplemen‐ tal oxygen (the latter by possibly masking hypercapnea and hypoventilation) are associated with a higher risk of unplanned cardiopulmonary events [9]. The monitoring period for the "event rate" should likely include the 30 days post procedure [10].

#### **2.1. Hypoxemia**

Hypoxemia, which is usually transient but often anxiety provoking for the colonoscopist, is a common occurrence during sedation for colonoscopy and has lead to the often "routine" practice by colonoscopists and attending anesthesiologists and nurse anesthetists of provid‐ ing patients with supplemental oxygen [9] delivered by nasal cannula and on occasion by a venturi air-entrapment mask (the latter providing a fixed and predictable oxygen concentra‐ tion despite a variable respiration pattern). Prolonged hypoxemia associated with colono‐ scopy is rare. The etiology of hypoxemia is often multifactorial but not to be overlooked is the amount of air (or carbon dioxide) insufflated into the colon for adequate luminal disten‐ tion and in some cases the passing of this gas into the small bowel through an incompetent ileocecal valve thereby affecting diaphragmatic function. This has lead most endoscopists to periodically monitor the degree of abdominal distention ("softness") by direct palpation of the abdomen either routinely during the procedure or when there are drops in the moni‐ tored oxygen saturation.

may reign supreme among the factors to consider. Intravenous replacement with crystalloid solutions should be considered during such events and in anticipation of such events in ap‐

The Major Complications of Colonoscopy: Sedation-Related, Hemorrhage Associated with Polypectomy and...

http://dx.doi.org/10.5772/51958

201

Cardiac arrhythmias including bradycardia, less often tachycardia, atrial premature contrac‐ tions, paroxysms of atrial fibrillation, and ventricular premature contractions have been documented during procedural sedation. Most resolve with the intravenous administration of fluids or increased sedation. With regard to bradycardia there is asymptomatic bradycar‐ dia (heart rate less than 60 bpm) and symptomatic bradycardia defined as a heart rate less than 60/min that elicits signs and symptoms. In symptomatic bradycardia the heart rate will usually be less than 50/min. Symptomatic bradycardia exists when the following 3 criteria are present: 1.) The heart rate is slow; 2.) The patient has symptoms; and 3.) The symptoms are due to the slow heart rate. Atropine is the first drug of choice for symptomatic bradycar‐ dia [12]. The dose in the bradycardia ACLS algorithm is 0.5 mg IV push which may be re‐ peated up to a total dose of 3 mg. (Anesthesiologists often choose glycopyrrolate [Robinul®] as an alternative to atropine.) Dopamine is a second line drug for symptomatic bradycardia when atropine is not effective. The dosage is 2-10 micrograms/kg/min infusion. Epinephrine can be used as an equal alternative to dopamine when atropine is not effective. The dosage is 2-10 micrograms/min. Rare cases of ventricular tachycardia and cardiac arrest during ven‐ tricular fibrillation have been reported necessitating the need for continuous EKG monitor‐ ing, the availability of ACLS trained personnel, as well as obtaining a history of cardiac or

Although the prevalence of coexistent pulmonary embolism at the time of colon cancer de‐ tection has been estimated to be as high as 2% (with the concurrent prevalence of deep ve‐ nous thrombosis being as high as 8%) there are no accepted statistics for the incidence of

The risk of stroke in patients with AF whose anticoagulation is adjusted for endoscopies is low, but almost tenfold higher in patients with complex clinical situations [14]. Age, history of stroke, hypertension, hyperlipidemia, and family history of vascular disease may increase the risk of suffering a stroke during or immediately after undergoing a gastrointestinal en‐ doscopic procedure. Comprehensive guidelines for the management of anticoagulation and antiplatelet therapy in patients undergoing gastrointestinal endoscopic procedures includ‐ ing colonoscopy have recently been published [15] and should serve as a reference and

pulmonary embolism complicating diagnostic or therapeutic colonoscopy [13].

propriately screened patients.

pulmonary disease prior to initiating the procedure.

**2.5. Pulmonary embolism**

guide when dealing with such patients.

**2.6. TIA/stroke**

**2.4. Arrhythmia**

#### **2.2. Hypercapnea**

Capnography monitor use is widespread in hospitals but these devices are less commonly used in the gastrointestinal endoscopy suite and other ambulatory settings where propofol is often used. In 2012 the American Society of Anesthesiologists (ASA) advised its member to use such carbon dioxide monitoring devices that detect changes in the amount of carbon dioxide the patient is exhaling during monitored anesthesia care for upper gastrointestinal endoscopies [11]. The new policy states:

*"Monitoring for exhaled carbon dioxide should be considered during endoscopic procedures in which sedation is provided with propofol alone or in combination with opioids and/or benzodiazepines, and especially during these procedures on the upper gastrointestinal tract. Careful attention to airway management must be provided during endoscopic retrograde cholangiopancreatography (ERCP) pro‐ cedures performed in the prone position where ventilatory monitoring, airway maintenance and re‐ suscitation may be especially difficult."*

It has only been a matter of time since this recommendation has found application in the monitoring of sedated patients undergoing colonoscopy. The new policy has met with mixed reviews in the gastrointestinal endoscopy and anesthesiology communities but most have agreed that there is little downside to such monitoring.

It is important to emphasize that at least one individual with training in advanced cardiac life support (tracheal intubation, defibrillation, use of resuscitation medications, ACLS cer‐ tification) that is capable of establishing an airway and providing positive-pressure ventila‐ tion should be present during colonoscopy sedation. Ability for communication with "back up" local paramedics or life support personnel should be confirmed.

#### **2.3. Hypotension**

The etiology of hypotension during colonoscopy is also multifactorial (pre-procedure anti‐ hypertensive medications, sedatives and analgesics used during the performance of the pro‐ cedure, arrhythmias, pre-procedure cardiac performance status, etc.) but the state of hydration of the patient after (usually a polyethylene glycol containing) bowel preparation may reign supreme among the factors to consider. Intravenous replacement with crystalloid solutions should be considered during such events and in anticipation of such events in ap‐ propriately screened patients.

#### **2.4. Arrhythmia**

practice by colonoscopists and attending anesthesiologists and nurse anesthetists of provid‐ ing patients with supplemental oxygen [9] delivered by nasal cannula and on occasion by a venturi air-entrapment mask (the latter providing a fixed and predictable oxygen concentra‐ tion despite a variable respiration pattern). Prolonged hypoxemia associated with colono‐ scopy is rare. The etiology of hypoxemia is often multifactorial but not to be overlooked is the amount of air (or carbon dioxide) insufflated into the colon for adequate luminal disten‐ tion and in some cases the passing of this gas into the small bowel through an incompetent ileocecal valve thereby affecting diaphragmatic function. This has lead most endoscopists to periodically monitor the degree of abdominal distention ("softness") by direct palpation of the abdomen either routinely during the procedure or when there are drops in the moni‐

Capnography monitor use is widespread in hospitals but these devices are less commonly used in the gastrointestinal endoscopy suite and other ambulatory settings where propofol is often used. In 2012 the American Society of Anesthesiologists (ASA) advised its member to use such carbon dioxide monitoring devices that detect changes in the amount of carbon dioxide the patient is exhaling during monitored anesthesia care for upper gastrointestinal

*"Monitoring for exhaled carbon dioxide should be considered during endoscopic procedures in which sedation is provided with propofol alone or in combination with opioids and/or benzodiazepines, and especially during these procedures on the upper gastrointestinal tract. Careful attention to airway management must be provided during endoscopic retrograde cholangiopancreatography (ERCP) pro‐ cedures performed in the prone position where ventilatory monitoring, airway maintenance and re‐*

It has only been a matter of time since this recommendation has found application in the monitoring of sedated patients undergoing colonoscopy. The new policy has met with mixed reviews in the gastrointestinal endoscopy and anesthesiology communities but most

It is important to emphasize that at least one individual with training in advanced cardiac life support (tracheal intubation, defibrillation, use of resuscitation medications, ACLS cer‐ tification) that is capable of establishing an airway and providing positive-pressure ventila‐ tion should be present during colonoscopy sedation. Ability for communication with "back

The etiology of hypotension during colonoscopy is also multifactorial (pre-procedure anti‐ hypertensive medications, sedatives and analgesics used during the performance of the pro‐ cedure, arrhythmias, pre-procedure cardiac performance status, etc.) but the state of hydration of the patient after (usually a polyethylene glycol containing) bowel preparation

tored oxygen saturation.

endoscopies [11]. The new policy states:

200 Colonoscopy and Colorectal Cancer Screening - Future Directions

*suscitation may be especially difficult."*

have agreed that there is little downside to such monitoring.

up" local paramedics or life support personnel should be confirmed.

**2.2. Hypercapnea**

**2.3. Hypotension**

Cardiac arrhythmias including bradycardia, less often tachycardia, atrial premature contrac‐ tions, paroxysms of atrial fibrillation, and ventricular premature contractions have been documented during procedural sedation. Most resolve with the intravenous administration of fluids or increased sedation. With regard to bradycardia there is asymptomatic bradycar‐ dia (heart rate less than 60 bpm) and symptomatic bradycardia defined as a heart rate less than 60/min that elicits signs and symptoms. In symptomatic bradycardia the heart rate will usually be less than 50/min. Symptomatic bradycardia exists when the following 3 criteria are present: 1.) The heart rate is slow; 2.) The patient has symptoms; and 3.) The symptoms are due to the slow heart rate. Atropine is the first drug of choice for symptomatic bradycar‐ dia [12]. The dose in the bradycardia ACLS algorithm is 0.5 mg IV push which may be re‐ peated up to a total dose of 3 mg. (Anesthesiologists often choose glycopyrrolate [Robinul®] as an alternative to atropine.) Dopamine is a second line drug for symptomatic bradycardia when atropine is not effective. The dosage is 2-10 micrograms/kg/min infusion. Epinephrine can be used as an equal alternative to dopamine when atropine is not effective. The dosage is 2-10 micrograms/min. Rare cases of ventricular tachycardia and cardiac arrest during ven‐ tricular fibrillation have been reported necessitating the need for continuous EKG monitor‐ ing, the availability of ACLS trained personnel, as well as obtaining a history of cardiac or pulmonary disease prior to initiating the procedure.

#### **2.5. Pulmonary embolism**

Although the prevalence of coexistent pulmonary embolism at the time of colon cancer de‐ tection has been estimated to be as high as 2% (with the concurrent prevalence of deep ve‐ nous thrombosis being as high as 8%) there are no accepted statistics for the incidence of pulmonary embolism complicating diagnostic or therapeutic colonoscopy [13].

#### **2.6. TIA/stroke**

The risk of stroke in patients with AF whose anticoagulation is adjusted for endoscopies is low, but almost tenfold higher in patients with complex clinical situations [14]. Age, history of stroke, hypertension, hyperlipidemia, and family history of vascular disease may increase the risk of suffering a stroke during or immediately after undergoing a gastrointestinal en‐ doscopic procedure. Comprehensive guidelines for the management of anticoagulation and antiplatelet therapy in patients undergoing gastrointestinal endoscopic procedures includ‐ ing colonoscopy have recently been published [15] and should serve as a reference and guide when dealing with such patients.

#### **2.7. Myocardial infarction**

Recent myocardial infarction has traditionally precluded the performance of elective colono‐ scopy for at least several months but recently this issue has been closely addressed in the literature [16]. Colonoscopy performed in patients who have experienced a recent myocar‐ dial infarction is associated with a higher rate of minor, transient, and primarily cardiovas‐ cular complications when compared with control patients but is relatively infrequently associated with major complications. In certain circumstances, despite the higher risk, colo‐ noscopy may be beneficial in this setting, particularly given the higher frequency of ische‐ mic colitis in this patient population. The occurrence of cardiac ischemia (and concomitant cardiac rhythm disturbances) in patients undergoing colonoscopy who have known cardiac disease or cardiac risk factors has recently been quantified [17]. Holter EKG recordings and measurement of cTnI troponin I levels showed a high incidence of new but silent ischemic and arrhythmic EKG changes during colonoscopy and do a lesser extent those patients with one or more risk factors for heart disease. Two patients with known heart disease died with‐ in 30 days of colonoscopy.

algorithm for such patients using these agents has recently been published [34]. Newer antico‐ agulants, for which current guidelines regarding their being held for endoscopic procedures, are lacking and these agents are reaching the market at an increasing rate. They include dana‐ paroid, a low molecular weight heparinoid consisting of a mixture of heparin sulfate, derma‐ tan sulfate, and chondroitin sulfate [35,36] which was recently removed from the US market due to shortages; the direct thrombin inhibitors recombinant hirudin (lepirudin), argatroban, desirudin and bivalirudin [37-39]; the recently available orally active direct thrombin inhibi‐ tor dabigatran etexlate [40]; and the factor Xa inhibitors idraparinux, rivaroxaban, and apixa‐ ban [41]. There is no doubt that as these agents are used they will affect practice standards with

The Major Complications of Colonoscopy: Sedation-Related, Hemorrhage Associated with Polypectomy and...

http://dx.doi.org/10.5772/51958

203

Acute postpolypectomy bleeding is often immediately localizable by or apparent on colono‐ scopy and amenable to endoscopic therapy using clips, ligatures, cautery or argon plasma coagulation [42,43] or nonendoscopic techniques such as angiographic embolizaton or sur‐ gery [44]. Recent endoscopic clip application devices have undergone redesign and im‐ provements to optimize their clinical effectiveness [45]. The site of delayed postpolypectomy colonic hemorrhage can be identified by colonoscopy, by red cell nuclear scintigraphy

A variety of procedural techniques have been proposed to minimize the risk of hemorrhage complicating polypectomy. These include the avoidance of the use of the "hot biopsy" tech‐ nique [47], the use of clips or detachable snares [48,49] and possibly the use of epinephrine injections to the base of the polyp prior to initiating the polypectomy [50,51]. Proper techni‐ que for the removal of pendunculated polyps includes planning for the application of pres‐ sure by regrasping the pedicle with the snare if immediate bleeding occurs, the injection of epinephrine 1:10,000 to 1:20,000 dilution into the bleeding site, the application of cautery (with thermal probes, bipolar cautery [BICAP], the argon plasma coagulator, or the tip of the polypectomy snare), the use of hemoclips, and/or the use of loops and band ligators on the pedicle. Similar techniques may be used in those with delayed bleeding who seem to be ac‐ tively bleeding. Up to 50 percent of patients with delayed hemorrhage may require blood

Perforation of the colon is the most dreaded complication of colonoscopy and polypectomy and this risk, albeit small, should be cited in the process of obtaining informed consent form

Abdominal pain, abdominal distention, +/- abdominal tenderness, hiccoughs, loss of bowel sounds indicative of ileus, and late developing peritoneal signs are the hallmarks of perfora‐ tion following colonoscopy. As physical examination, chest x-ray and abdominal flat and upright x-ray alone or in combination may not be diagnostic of colonic perforation patients

and/or by selective angiography [46] and dealt with in a similar fashion.

**4. Colonic perforation associated with polypectomy**

regard to colonoscopy and polypectomy.

transfusion [23].

the patient for the procedure.

### **3. Colonic hemorrhage associated with polypectomy**

Although hemorrhage can occur during diagnostic colonoscopy (particularly when "cold" or "hot" forceps biopsy techniques are used) it most often complicates polypectomy occur‐ ring immediately or being delayed for several weeks after the procedure [18]. The overall incidence of hemorrhage has been reported to be in the range of 1 to 6 per 1,000 colonoscop‐ ies [19,6] (this being a useful figure to quote when obtaining informed consent from the pa‐ tient) with the number of polyps [20], polyp size [19] recent anticoagulant use [22,23] and even polyp histology(!) [6,24,25] being modifying factors! The effects of aspirin, nonsteroidal anti-inflammatory drugs and clopidogrel both alone and in combination on this complica‐ tion of colonoscopy have also been addressed [26,27].

Patients requiring colonoscopy with or without biopsy and/or polypectomy are often taking antithrombotic agents including anticoagulants such as warfarin, heparin, and low molecular weight heparin, and antiplatelet agents such as aspirin, non-steroidal anti-inflammatory drugs, thienopyridines such as clopidrogel and ticlopidine, and glycoprotein IIb/IIIa receptor inhibitors. The indications for the use of these medications include atrial fibrillation, acute cor‐ onary syndrome, deep venous thrombosis, hypercoagulable states and indwelling endopros‐ theses such as coronary artery stents. When hemorrhage does occur in patients taking these agents it is most commonly from the gastrointestinal tract [28]. Risk stratification for these pa‐ tients can generally be relegated to two categories. Low risk procedures include diagnostic co‐ lonoscopy including mucosal biopsy [29,30] and high-risk procedures include colonoscopy with polypectomy and the dilatation of either benign or malignant colonic strictures (guide‐ lines extrapolated in part from experience reported in the upper gastrointestinal endoscopy lit‐ erature) [31-33]. A comprehensive review of the types of antithrombotic therapies, their implications for patients undergoing colonoscopy, and recommendations and a management algorithm for such patients using these agents has recently been published [34]. Newer antico‐ agulants, for which current guidelines regarding their being held for endoscopic procedures, are lacking and these agents are reaching the market at an increasing rate. They include dana‐ paroid, a low molecular weight heparinoid consisting of a mixture of heparin sulfate, derma‐ tan sulfate, and chondroitin sulfate [35,36] which was recently removed from the US market due to shortages; the direct thrombin inhibitors recombinant hirudin (lepirudin), argatroban, desirudin and bivalirudin [37-39]; the recently available orally active direct thrombin inhibi‐ tor dabigatran etexlate [40]; and the factor Xa inhibitors idraparinux, rivaroxaban, and apixa‐ ban [41]. There is no doubt that as these agents are used they will affect practice standards with regard to colonoscopy and polypectomy.

**2.7. Myocardial infarction**

202 Colonoscopy and Colorectal Cancer Screening - Future Directions

in 30 days of colonoscopy.

**3. Colonic hemorrhage associated with polypectomy**

tion of colonoscopy have also been addressed [26,27].

Recent myocardial infarction has traditionally precluded the performance of elective colono‐ scopy for at least several months but recently this issue has been closely addressed in the literature [16]. Colonoscopy performed in patients who have experienced a recent myocar‐ dial infarction is associated with a higher rate of minor, transient, and primarily cardiovas‐ cular complications when compared with control patients but is relatively infrequently associated with major complications. In certain circumstances, despite the higher risk, colo‐ noscopy may be beneficial in this setting, particularly given the higher frequency of ische‐ mic colitis in this patient population. The occurrence of cardiac ischemia (and concomitant cardiac rhythm disturbances) in patients undergoing colonoscopy who have known cardiac disease or cardiac risk factors has recently been quantified [17]. Holter EKG recordings and measurement of cTnI troponin I levels showed a high incidence of new but silent ischemic and arrhythmic EKG changes during colonoscopy and do a lesser extent those patients with one or more risk factors for heart disease. Two patients with known heart disease died with‐

Although hemorrhage can occur during diagnostic colonoscopy (particularly when "cold" or "hot" forceps biopsy techniques are used) it most often complicates polypectomy occur‐ ring immediately or being delayed for several weeks after the procedure [18]. The overall incidence of hemorrhage has been reported to be in the range of 1 to 6 per 1,000 colonoscop‐ ies [19,6] (this being a useful figure to quote when obtaining informed consent from the pa‐ tient) with the number of polyps [20], polyp size [19] recent anticoagulant use [22,23] and even polyp histology(!) [6,24,25] being modifying factors! The effects of aspirin, nonsteroidal anti-inflammatory drugs and clopidogrel both alone and in combination on this complica‐

Patients requiring colonoscopy with or without biopsy and/or polypectomy are often taking antithrombotic agents including anticoagulants such as warfarin, heparin, and low molecular weight heparin, and antiplatelet agents such as aspirin, non-steroidal anti-inflammatory drugs, thienopyridines such as clopidrogel and ticlopidine, and glycoprotein IIb/IIIa receptor inhibitors. The indications for the use of these medications include atrial fibrillation, acute cor‐ onary syndrome, deep venous thrombosis, hypercoagulable states and indwelling endopros‐ theses such as coronary artery stents. When hemorrhage does occur in patients taking these agents it is most commonly from the gastrointestinal tract [28]. Risk stratification for these pa‐ tients can generally be relegated to two categories. Low risk procedures include diagnostic co‐ lonoscopy including mucosal biopsy [29,30] and high-risk procedures include colonoscopy with polypectomy and the dilatation of either benign or malignant colonic strictures (guide‐ lines extrapolated in part from experience reported in the upper gastrointestinal endoscopy lit‐ erature) [31-33]. A comprehensive review of the types of antithrombotic therapies, their implications for patients undergoing colonoscopy, and recommendations and a management Acute postpolypectomy bleeding is often immediately localizable by or apparent on colono‐ scopy and amenable to endoscopic therapy using clips, ligatures, cautery or argon plasma coagulation [42,43] or nonendoscopic techniques such as angiographic embolizaton or sur‐ gery [44]. Recent endoscopic clip application devices have undergone redesign and im‐ provements to optimize their clinical effectiveness [45]. The site of delayed postpolypectomy colonic hemorrhage can be identified by colonoscopy, by red cell nuclear scintigraphy and/or by selective angiography [46] and dealt with in a similar fashion.

A variety of procedural techniques have been proposed to minimize the risk of hemorrhage complicating polypectomy. These include the avoidance of the use of the "hot biopsy" tech‐ nique [47], the use of clips or detachable snares [48,49] and possibly the use of epinephrine injections to the base of the polyp prior to initiating the polypectomy [50,51]. Proper techni‐ que for the removal of pendunculated polyps includes planning for the application of pres‐ sure by regrasping the pedicle with the snare if immediate bleeding occurs, the injection of epinephrine 1:10,000 to 1:20,000 dilution into the bleeding site, the application of cautery (with thermal probes, bipolar cautery [BICAP], the argon plasma coagulator, or the tip of the polypectomy snare), the use of hemoclips, and/or the use of loops and band ligators on the pedicle. Similar techniques may be used in those with delayed bleeding who seem to be ac‐ tively bleeding. Up to 50 percent of patients with delayed hemorrhage may require blood transfusion [23].
