Hi, I'm Doug Rex from Indiana University Hospital in Indianapolis, and I want to welcome you to a discussion of colonoscopy technique. This video is sponsored by the American Society for Gastrointestinal Endoscopy. The content of the video includes a discussion of informed consent, selecting the type of colonoscope, understanding tip deflection, anatomic landmarks in the colon, colonoscopic insertion technique, intubation of the terminal ileum, impact of sedation on technique, maintaining safety during colonoscope insertion, withdrawal technique, and finally, the very difficult colonoscopy. You can view these sections in sequence or select from this menu. A separate DVD covers polypectomy technique. In these discussions, I'm going to rely heavily on my personal experience performing colonoscopy and on the teachings of many well-known colonoscopists. On my personal experience with medical legal cases involving colonoscopy, and I'll refer to and defer to the published medical literature when possible. Let's get started. The first important question about informed consent is who should obtain it. A study of time utilization in Indiana and experience in medical legal cases indicate that there is variability across the U.S. with regard to who obtains the consent. In many practices, the endoscopist obtains consent, but in some practices, a nurse or other medical professional obtains consent. Therefore, there is no standard of medical care with regard to who obtains consent. My opinion, based on review of medical legal cases, is that the best situation for defense of a medical legal action is when the endoscopist has obtained consent and the patient remembers a clear-cut discussion of the risks with the endoscopist. It can also be useful to provide written material or a video for the patient to learn about the procedure and its risks. The issue of when consent should be obtained pertains to whether consent can be obtained in the open access process on the day of the procedure. There's a trend of regional variation in the U.S. regarding whether endoscopy units use closed access or open access. Closed access refers to when patients are routinely seen in consultation by the endoscopist prior to and on a different day than when colonoscopy is performed. Closed access is relatively common on the coasts of the United States, whereas open access is more common in the central part of the country. Open access means that some patients are scheduled for colonoscopy and appear prepped in the endoscopy unit without ever having met the colonoscopist. In this setting, informed consent will be obtained by the endoscopist immediately prior to the procedure. Open access is sufficiently common in the U.S. that we must conclude that the standard of medical care allows for informed consent to be obtained on the day of the procedure. The specific risks that should be mentioned include perforation, along with the likely need for surgery if perforation occurs, the possibility of a missed lesion or cancer, bleeding, and medication reaction. Colonoscopists should consider discussing the risk of splenic injury. From a medical legal perspective, perforation and the possibility of missed lesions should be emphasized since these are the two complications most likely to result in a malpractice suit. Splenic injury may result in a malpractice suit if its management requires splenectomy. If you know your own perforation rate, you can quote it. If you don't know your own perforation rate, studies show the overall perforation rate in the United States Medicare population for colonoscopy is still about 1 in 500, although it's somewhat less than 1 in 1,000 for screening cases. Many experienced colonoscopists have a perforation risk of 1 in several thousand for screening colonoscopy. Screening means doing colonoscopy in asymptomatic people to look for neoplasia, and asymptomatic people are less likely to have an abnormal colon that would increase the risk of perforation. A major deficiency of colonoscopy as a diagnostic, surveillance, and screening test is that the procedure is highly operator-dependent. Several studies have shown that general surgeons and primary care physicians performing colonoscopy are more likely to fail to detect colorectal cancer than our gastroenterologists. This difference suggests, but does not prove, that training and detection is associated with better protection against colorectal cancer. The specific elements of training that lead to better detection have not yet been defined. However, even within members of the same gastroenterology group, it's been repeatedly shown that detection of adenomas is highly variable. The next slide shows that detection of proximal serrated lesions is even more variable. This means that some gastroenterologists miss the majority of adenomas in the colon during colonoscopy, miss the majority of advanced adenomas, miss the majority of serrated lesions in the colon also. Since 2002, guidelines have recommended that colonoscopists measure the quality of mucosal inspection by measuring the adenoma detection rate. The adenoma detection rate is defined as the percentage of patients age 50 and older undergoing their first screening colonoscopy who have one or more adenomas detected. The recommended threshold for adequate detection is at least 25% of men and at least 15% of women should have one or more adenoma detected. The true prevalence of adenomas in the population is considerably higher than these thresholds, as demonstrated by repeated studies showing overall adenoma detection rates in the 40 and even 50% and higher range using high-definition colonoscopes. In a study from Poland of screening colonoscopy, doctors who had adenoma detection rates below the recommended threshold for a mixed-gender population of 20% saw their patients have hazard ratios for the development of cancer that were more than 10 times those of patients who had undergone colonoscopy by doctors with adenoma detection rates above 20%. Thus, measuring the adenoma detection rate is a must for every colonoscopist and is a validated measure of the quality of mucosal inspection. Several recent studies have demonstrated that colonoscopy is less effective at preventing cancer on the right side of the colon compared to the left side of the colon. In the best studies, protection rates on the right side are in the range of 50-60% while they exceed 80% in the distal colon. Several reasons have been proposed for why right-sided protection might be worse. First, some doctors have lower rates of cecal intubation, and some may say they have achieved cecal intubation when they have not. Thus, current guidelines for colonoscopy quality recommend minimum thresholds for cecal intubation and recommend documentation of cecal intubation by notation of landmarks in the cecum and photography of visualized landmarks. Second, bowel preparation tends to be worse on the right side of the colon than the left side. This discrepancy in bowel preparation can be overcome by using either split dosing or same-day dosing of bowel preparations. Split dosing refers to giving half the preparation the evening before colonoscopy and the other half on the morning of colonoscopy. Split dosing counters the tendency of the right colon to have adherent mucus and small intestinal secretions when the interval between completing the preparation and performing the colonoscopy is too long. For patients scheduled for afternoon colonoscopies, ingestion of the entire preparation on the morning of the examination is associated with better quality preparation and may be more convenient for patients than split dosing. A third potential explanation for less protection on the right side of the colon is that lesions that are endoscopically subtle, including conventional adenomas that are flat or depressed, as well as serrated lesions, including hyperplastic polyps and sessile serrated polyps, are more common in the proximal colon and may be missed more often. Thus, the modern colonoscopist must be trained in the recognition of the full spectrum of precancerous lesions in the colon. Finally, the biology of colorectal cancer is different in right-sided compared to left-sided cancers, with more right-sided cancers demonstrating hypermethylation and or microsatellite instability. This may result, in some instances, in a more rapid conversion through the polyp cancer sequence. In other words, a different biology in right-sided cancers compared to left-sided lesions. An important emphasis of this video and the accompanying video on polypectomy will be colonoscopic technique and technologies that minimize the chance of a missed lesion. Nevertheless, in the informed consent process, missing must be discussed because the consequences of developing cancer shortly after colonoscopy may be very serious for the patient. And whether such a cancer was the result of missing or not, it may result in medical legal action. The largest tandem colonoscopy study found that the miss rate for adenomas 1 cm and larger was 6%, and two CT colonography studies from Centers of Excellence have measured the miss rate of optical colonoscopy for adenomas 1 cm and larger at 12% and 17%. Endoscopists typically do not know their miss rates. They should be able to respond to patient queries about their skills in mucosal detection by giving their personal adenoma detection rates. It is perfectly reasonable for a patient contemplating colonoscopy to ask their potential colonoscopist for their adenoma detection rate. There also should be a discussion of the risk of bleeding, which is essentially confined to polypectomy, reaction to the sedative medications used, and there should be an opportunity to ask questions. Here's the whole process as I do it. So we've decided that we need to do a colonoscopy. It's indicated, but I want to make sure that you understand that there are some risks to the procedure. So first of all, we're going to sedate you. We want you to be comfortable. The risks of that include a small chance of dropping your blood pressure or having a change in your vital signs, including not breathing deeply enough. We're going to be monitoring all of that very closely. We're experts at taking care of any problems that might arise. Also because you'll be sleepy, there's a slight risk of getting something down your windpipe, what we call aspiration, which might result in bronchitis or pneumonia. That chance is very small, less than one in a thousand. The risk of perforation, making a hole in the colon, it's a very serious thing. It can be bad enough to require surgery to fix it, is one in several thousand or less. If I take polyps out of your colon, these little growths that can turn into cancer, there's a chance of having bleeding. It's primarily present if we run into a large polyp. Those, of course, are the ones that are most important to remove. That can happen any time in the first couple of weeks after the procedure. After the procedure, I'll be able to give you an idea of whether I think there's any significant risk of that based on whether we find polyps and how large they are. We say that colonoscopy is the best test available to look at the colon, but it's important to understand that it's not perfect. Rarely, it can miss something, including something significant. Finally, it's possible but very rare for colonoscopy to injure the spleen. There have been reports of injury that's been severe enough that it actually required surgical removal of the spleen. I think the chances of that are extremely small, but I just need for you to know that. Any questions for me? I don't think so. An injury that was seldom recognized during colonoscopy two decades ago, but is now seen with increasing frequency, is splenic injury. The mechanism by which splenic injury occurs during colonoscopy is uncertain, but it is assumed that complex loop formation during colonoscopy results in tension on the splenocolic ligament that leads to laceration or avulsion of the capsule with bleeding from the spleen surface or laceration with bleeding or formation of an intrasplenic hematoma. Diagnosis is often delayed and typically presents with left-sided abdominal pain, sometimes pain in the left shoulder, known as care sign, and can present with shock. Patients who are hemodynamically unstable may require splenectomy, and these patients may bring a malpractice action. Risk factors for splenic injury include female gender and possibly factors such as deep sedation, cigarette smoking, and prior abdominal surgery with adhesion formation around the spleen. However, only female gender is established as a definite risk factor, with about three quarters of reported cases occurring in women. There is no established rule for how to avoid splenic injury during colonoscopy. My anecdotal recommendation is to be cautious and gentle with application of torque when the colonoscope tip is proximal to the splenic flexure and there are loops in the instrument. By insertion tube selection, I am referring to the choice between an adult or standard diameter insertion tube and the narrower pediatric insertion tube. For routine cases, the selection between these devices is, I believe, purely a matter of personal preference. The standard insertion tube has the advantage of being thicker and stiffer and therefore less likely to cause looping during insertion, particularly in the transverse colon. In my experience, standard insertion tubes are associated with about a 20% reduction in the need for ancillary maneuvers such as abdominal pressure during insertion. They also have larger working channels, typically at least 3.7 millimeters, and therefore suctioning is faster. Suctioning can continue when there are devices in the insertion tube and they are better able to suck polyps back through the channel and clots in the case of lower gastrointestinal bleeding. The thinner pediatric insertion tube is more flexible and therefore, on average, will result in less pain because it generates less force in stretching the mesentery. The super or ultra-thin colonoscopes have diameters that are similar to upper endoscopes and may further decrease the tendency toward pain during insertion relative to pediatric colonoscopes. You may be aware of studies showing that flexible sigmoidoscopy pain can be reduced by using thin upper endoscopes to perform flexible sigmoidoscopy. Thus, upper endoscopes, which are sometimes used to perform colonoscopy in patients with angulated sigmoid colons, are even thinner than pediatric colonoscopes, generate very little force on the mesentery, and therefore produce less pain. Pediatric scopes are also very useful in passing strictures. About half the market in colonoscopes in the US is for pediatric scopes, indicating that many colonoscopists prefer pediatric scopes for routine use. In passing a stricture or sharp angulation, it should be remembered that even a small reduction in diameter can produce a substantial reduction in the cross-sectional surface area of the colonoscope tip, which is the critical determinant of ability to pass the stricture or angulation. Thus, moving from a standard insertion tube to a pediatric insertion tube is associated with a 1.5 millimeter reduction in diameter, but a 22% reduction in surface area. Moving from a pediatric colonoscope to an upper endoscope is associated with another 2.7 millimeter reduction in diameter and a 42% reduction in surface area. Finally, a thinner insertion tube can be successful when a larger insertion tube has been unsuccessful because of fixation. Fixation is the sensation that the colonoscope tube cannot be advanced because the colon is stuck in a fixed position. Normally, the colon can be felt to move and the bowel stretches on its mesentery with advancement or loop formation. This is something that is palpated by the hand that is advancing the colonoscope. If the colonoscope is felt to meet fixed resistance, then insertion must stop, even if the lumen of the colon is clearly in view. In this instance, it may be of value to switch to a thinner endoscope because the increased flexibility and reduction of force applied to the colon wall may allow the thinner instrument to be passed further without meeting fixed resistance. Having made these comments about the relative thickness and flexibility of the standard and pediatric insertion tubes, certain commercial colonoscopes are now available with a variable stiffness function. Using a dial on the control head, the stiffness of the insertion tube can be varied from a stiffness that approximates that of the most flexible pediatric colonoscopes to the stiffness that approaches the range of the stiffest standard insertion tubes that have been commercially available. Variable stiffness is most often employed as a method to reduce looping in the sigmoid after the sigmoid has been passed by the colonoscope tip. Thus, insertion begins with the stiffness device set on zero, or the pediatric colonoscope flexibility range, and the sigmoid colon is negotiated. After the colonoscope tip has been advanced beyond the sigmoid colon, the instrument is straightened and the device is activated so that the stiffening of the insertion tube may reduce the tendency of the colonoscope to form sigmoid loops. Many colonoscopists, myself included, like to have variable stiffness available and feel that in specific instances it seems to help colonoscope tip advancement. Here are some general observations about variable stiffness that may be of value to you. First, you are more likely to find the variable stiffness function valuable when using pediatric colonoscopes rather than standard insertion tubes. With standard insertion tubes, there are a greater percentage of cases where the tip can be readily advanced to the cecum without activating the stiffener, just because the thicker colonoscope tends to reduce loop formation in any case. Second, there will be many instances where activation of the stiffener seems to make no difference in colonoscope tip advancement. Therefore, colonoscopists must be skilled in the other ancillary maneuvers that resist loop formation, including position change and abdominal pressure. Third, I am convinced that there are specific instances in which activation of the stiffener actually impedes advancement of the colonoscope tip. Therefore, if you have activated it to help pass an area of difficulty, and now you encounter another area of difficulty later in the examination and the stiffener is still activated, remember to try to deactivate the stiffener and place the colonoscope in the more flexible mode again. The data on the effectiveness of variable stiffness can be summarized as follows. There is no evidence that variable stiffness improves the success rate of sacral intubation. The data on reduction in sacral intubation times is quite mixed, but there probably is some improvement in sacral intubation time, particularly for trainees and people early in their experience of colonoscopy. There is also no clear-cut evidence of improved sacral intubation rates in patients with prior failed colonoscopies. A number of studies do show a reduction in the need for ancillary maneuvers, typically on the order of 10-15%, and particularly in the need for abdominal pressure. If I could summarize the situation, I would say that lots of people like to have variable stiffness available and believe that it helps, but the hard evidence suggests that there is only a mild improvement in colonoscopy insertion associated with variable stiffness. A very recently marketed development in insertion tubes has been called passive bending. This refers to a section of the insertion tube immediately adjacent to the bending section that has the property of distributing force over a longer section of the colonoscope when passing turns. The value of this passive bending development is not yet certain, but its presence is not expected to require fundamental changes in insertion technique. Returning to the issue of selecting the standard versus pediatric colonoscope, I mentioned that for routine cases, this selection is a matter of personal preference of the endoscopist. However, an effective colonoscopist must have a pediatric colonoscope available and be willing to use it in selected circumstances. The pedoscope should be selected in patients who have known severe diverticular disease, known severe angulation in the sigmoid, or known fixation of the sigmoid, which sometimes occurs as a result of prior pelvic or abdominal surgery. The pediatric instrument may be a superior choice when unsedated colonoscopy is considered particularly in women because the increased flexibility reduces the pain from colonoscope passage. I prefer the pediatric colonoscope for all cases of Crohn's disease because of the ability to more easily pass strictures and enter the terminal ilium. The pediatric colonoscope is a wise choice when you have to get into the terminal ilium, for example, in a patient with unexplained abdominal pain and diarrhea because it is just easier to enter the TI with the pediatric scope. I have a personal preference for the pediatric colonoscope for some cases of large sessile polyp resection because of increased maneuverability of the tip, including the ability to retroflex more easily in the cecum, the ascending colon, and the transverse colon. The pediatric colonoscope should also be selected when you know there is an increased risk of mechanical perforation. I include here radiation colitis, patients on chronic corticosteroids, and those with known colonic amyloidosis. The colon is more fragile in these settings, and the reduced forces associated with a pediatric colonoscope will likely reduce the risk of mechanical perforation. In fact, if you are using a pediatric variable stiffness colonoscope, I recommend not activating the stiffener when colonoscopy is performed for these indications. Before we begin colonoscopy, we must understand the operational features of the bending section, which directs the tip of the scope through the colon. Commercial colonoscopes essentially all have the same deflection properties when working properly and will allow 180 degrees of bending in the up direction and 180 degrees in the down direction. Right-left maximum deflection is 160 degrees in both directions. The control knob that is easiest to access is the up-down control. The bending head can easily be deflected in the up direction by using the thumb, but it is most easily directed in the down deflection using a combination of the thumb and the fingers which pull the control knob down. I believe this is also the most ergonomically sound way to manipulate the up-down control. Many colonoscopists manipulate the right-left control with their thumb also. My opinion is that for many of us, manipulation of the right-left control with the thumb is not an ergonomically sound maneuver. I prefer to take my right hand off the instrument shaft to operate the right-left control. There are two ways during colonoscopy to achieve right-left movement of the scope. One of these is by torque. In this maneuver, the colonoscope is deviated in the up-down direction, usually the up direction, and the instrument shaft is torqued with the right hand. This is a fundamental movement during both insertion and withdrawal of the scope. The second method, of course, is to deviate the tip using the right-left control. During insertion of the instrument, pain is related primarily to loop formation, but in some persons with sensitive or angulated colons, significant pain can be induced by torquing the instrument. Therefore, when pain during insertion is a problem, such as in the unsedated or lightly sedated patient, right-left deviation by use of the control knob may be more comfortable for the patient than right-left deviation achieved by torquing. True precision control of tip movement is best achieved by simultaneously controlling the up-down knob with the thumb and fingers of the left hand and the right-left knob using the right hand. In order to employ this two-handed control of the tip direction, we will need to learn what I refer to as the left-handed scope grip. Here is an example of a very difficult turn in the sigmoid colon that requires precise deflection of the instrument tip. In order to achieve this, we have placed the instrument shaft between the fourth and fifth fingers of the left hand. This maneuver allows us to continue to place forward force on the colonoscope using the left hand while simultaneously using the left hand to control the up-down control and getting precise tip deflection by using the right hand to control the right-left knob. You'll see me use this grip to achieve precise tip control during insertion and in a separate DVD I'll use it to achieve exquisite precision during the performance of complex polypectomy. The advantage of this grip, which I like to call the left-handed scope grip, is that it will allow us to achieve simultaneous insertion or withdrawal and precision control of the up-down and right-left knobs. There is one additional point that I would like to make about tip deflection that is very important during colonoscopy. We discussed earlier that we can deflect the tip 180 degrees in the up-down direction and 160 degrees in the right-left direction. However, if we initially deflect the tip maximally in the up-down direction and then maximally deflect the tip in the right-left direction, the scope tip will hairpin. That is, it will look back at itself, but it will deviate very little from the up-down plane. This can be extremely useful in passing very sharp turns. With the scope passing through a turn in the maximum 180 degree up direction, the lens is against the wall. By deflecting maximum left in addition to maximum up, the tip now turns into the lumen, completing passage of the turn. Here is the real-time demonstration. We see the colonoscope tip passing through a turn in the sigmoid colon using what is known as the slide-by maneuver, which will be discussed later. Notice that the up-down control has been deflected maximally in the up direction, but by stopping and using the left-hand shaft grip and turning the right-left control toward the left, the lumen is brought into view. This hairpin maneuver is critical during retroflexed examination, and as we just discussed, it can be extremely useful in passing sharp turns during insertion. It's useful for a colonoscopist to know the standard information about landmarks, but I want to say at the outset that there are only two areas of the colon where you can reliably say that you know for sure where you are, i.e. what section of the large bowel you are in. One of these areas is the rectum and the distal sigmoid. The rectum can be recognized by its large luminal diameter and its distinct series of valves, but the most important indicator of the rectum is the distance from the external verge. The proximal margin of the rectum in adults is reliably about 15 cm from the external verge. You can pretty reliably conclude that you are at the rectosigmoid junction at around 15 cm, and in the distal sigmoid at around 17-18 cm, up to about 25-30 cm. Once the scope has been inserted beyond 30-35 cm, there's going to be uncertainty as to precisely what section of the colon the scope tip is in until you reach the cecum. The other area of the colon that we can identify with absolute certainty is the cecum because no other area of the GI tract has the appearance of the cecum. To be certain we're in the cecum, we must see the ileocecal valve and its orifice, and we should also see the appendiceal orifice. Only occasionally is the appendiceal orifice completely obliterated by prior appendectomy. Intubation of the terminal ileum can help assist identification of the cecum, but it's not necessary for reliable identification. We can also be reasonably certain that we are in the ascending colon if the ileocecal valve is still visible to us in the endoscopic field of view. When we are outside of these regions, lesions that are sent for surgical resection, particularly laparoscopic resection, or lesions that must be identified on follow-up endoscopic examination should be tattooed with carbon black. General clues to location include the following. The sigmoid tends to be tubular, narrow relative to the more proximal colon, to have multiple angulations on insertion, in particular compared to the descending colon, which tends to be straight, and in some patients extensive diverticular disease is a clue to the sigmoid. With the patient in the left lateral decubitus position, the splenic flexure is often marked by a pool of fluid and a sharp angle. The transverse colon tends to be triangular, and because of its anterior location, light transmission through the abdominal wall and light moving across the upper abdomen with scope movement is a good clue. The hepatic flexure is often notable for the bluish color created by the liver on the opposite side of the bowel wall and its location adjacent to the ascending colon. The standard of medical care requires that a colonoscopist who sends a patient for surgical resection of a lesion accurately locate that lesion for the operating surgeon. Remember to never get cocky about where you think a lesion is located. If it's not in the cecum or in the right colon near the ileocecal valve, tattoo it. In the rectum and sigmoid, even though you may be confident that you know the section of bowel that the lesion is in, it's still often best to tattoo the lesion to make it very clear to the surgeon. Tattooing the distal end of a rectal lesion can help the surgeon to understand how distal their dissection must be carried. So we've done the history and the examination, the patient's given informed consent, we're ready to start the procedure. The first step in the procedure is always a careful rectal examination. So we're going to use a lot of lubricant. We want to do a quick inspection of the outside of the anus and the rectum and then lube up, palpate the anal canal and then do a full circular examination of the rectum. It's particularly important to examine the posterior fornix of the rectum. When the scope is inserted, it is directed toward the anterior wall and the posterior wall in a very large rectum may not be seen well. A prostate exam should be performed and the rectal examination results should be documented in the colonoscopy report. Now we're ready to insert the scope. So we've got a lot of lubricant on the scope. We usually want to use our index finger to direct the scope up to the anus and we want to point it sort of toward the umbilicus. During the rectal examination, we really can get a sense of the angle that the anal canal takes. And we're going to sort of remember that angle and follow it with the scope. So in order to introduce the scope, we put the edge of it into the anus and then turn it in the direction of the canal and insert it and then we're ready to go. Now there's two approaches to where to put the scope during the examination, especially the beginning part of the examination where we still have a lot of the scope out of the patient. A very common approach is to lay the scope up on the examining table. And an advantage of this is that it's very stable. It's not going to fall out and fall onto the floor, which could damage the scope. So some people like this. My own tendency is to let the scope hang over the edge and then if I need to take my hand off the scope, I simply pin it against the side of the bed with my hip. And that will keep the scope from falling out of the patient. Again, once we have a lot of the scope out of the patient, the simple weight of the scope could pull it out of the patient if we don't pin it in some place. My own preference for this technique is basically that when the scope is lying up on the bed, if you torque the instrument or turn it, you can feel two torsion points. One is inside the patient and the second one is this point right here that's on the bed. And I would prefer to not sense that and to only feel what's going on with the scope inside the patient. So I like to let it hang free and now everything that goes on with the scope and that I feel in my hand as I maneuver the scope will reflect something that's going on inside the patient. Next, we want the colon to be wet during insertion. When we enter the colon, we may see a range of wetness that goes from pools of fluid on the one hand to a colon that is so dry at the other extreme that the mucosa is literally peeling apart as we insufflate air. A wet colon improves the hydrophilic forces between the colonoscope and the colon wall and facilitates insertion. If the colon appears at all dry, we should add water, which we can do by either pushing a syringe full of water down the biopsy channel or by using the water jet available now in many commercial colonoscopes in which pushing a foot pedal sends a stream of water into the colon. Next, we'll demonstrate how to determine the direction of the lumen during colonoscope insertion. For this and some other sections of the video, we'll use the Olympus scope guide to facilitate our demonstrations. This device is a magnetic electronic imager that displays the shape of the colonoscope shaft within the patient and helps us to understand loop formation and scope manipulations. In this picture, we see the endoscopic view and below it the shape of the colonoscope within the patient as demonstrated by the scope guide. Three visual cues can point to the direction of the lumen. These include first, the curves formed by haustral and other colonic folds and sometimes curves are also formed by the light reflection, or the so-called highlights, second, the shadowing produced by folds, and third, the direction of the tineal folds. The tinea extend from the appendiceal orifice to the rectosigmoid junction and consist of the longitudinal muscle layer of the colon wall, which is grouped into three bundles called tinea. As we will see, these tineal folds create an impression or indentation that is often visible in the colonic lumen, particularly in the cecum and in the flexures and sharp angulated bends proximal to the sigmoid. As we travel through the colon, we will invariably have visual cues that point to the direction of the lumen unless the scope tip abuts against or is too close to the mucosa. For example, if any semicircular colonic haustral fold is in the endoscopic view, the lumen will be in the direction of the middle or center of a circle formed by the curve of the fold. The highlights or light reflections may also form a semicircle whose concave side points to the direction of the lumen. If a haustral fold casts a shadow on the colonic wall in the distance, that shadow is on the side in which the lumen is turning proximal to the haustral fold. Tineal folds produce a linear impression on the colonic lumen that points to the direction the colon is turning. Even at very close proximity of the scope tip to the colon wall, small portions of the colonic folds or relatively very flat folds may be visible and will point to the direction of the lumen, such as here. Here is an example of how even a relatively flat fold with minimal curvature still indicates reliably that the lumen is in the up direction. Here again, the circular folds point reliably to the direction of the lumen. In this view, both the shadow created by the haustral fold and the prominent tineal fold point to the luminal direction. Note that even this relatively flat fold and the surrounding highlights point accurately to the luminal direction. As we enter the cecum, we see that the convex side of the curves formed by the haustral folds continue to point in the direction of the lumen. Another technique we need to learn is the slide-by maneuver. At times, the colon will bend so sharply that we cannot keep the lumen in view. In that case, we have to determine the direction of the lumen and then gently advance the scope through the turn without direct vision. Thus, the lumen will not be in view as we pass through the turn and we will have a so-called red-out appearance when the scope tip is up against the mucosa. As long as we see the mucosa sliding by us in the direction opposite the one we are headed, we're okay and we can continue to advance through the turn. Here is a demonstration of a slide-by. We have entered the sigmoid and we see a sharp turn in the up direction. Deflecting the tip now in the up direction, we see the lumen sliding past us in the down direction. We're jiggling the scope a bit in order to help the scope tip slide through the turn and we are able to complete passage of the turn. Here is another example of the slide-by maneuver with the luminal direction in the up left and we see the mucosa sliding by us in the down and right direction as we pass through the turn. Now let's put these elements of luminal direction and slide-by together during a secal insertion. In this view, we see the colonic folds pointing in the up direction to show us the lumen. As we push in the direction indicated by the circular folds, we see the mucosa slipping past in the opposite direction in the slide-by maneuver. Shadowing caused by the haustral fold in the down right direction indicates that the lumen will turn in that direction. Here we see the circular folds pointing to the lumen in the right direction as we slide through the turn. As we follow the insertion, we will repeatedly recognize how the circular folds and shadowing point in the direction of the lumen. As we proceed into the proximal colon, the tinea impression will also help designate the luminal direction. As we move forward, you'll recognize that clues to the luminal direction are almost continuously in view, even though the scope tip is progressing rapidly, and they continue right into the sacral tip. Now that we can recognize the lumen and control tip deflection, how do we get the scope to the cecum? It's certainly not by just continuing to push in the direction of the lumen, because this will inevitably lead to the formation of bends or twists in the instrument shaft, typically referred to as loops in colonoscopy. Loops are counterproductive because they are the main cause of pain during insertion, and because any time a loop or bend forms, additional pushing on the instrument shaft will tend to transmit force into the bend or loop itself, and thereby increase its size, as opposed to transmitting force to the tip of the colonoscope and advancing the instrument tip. Therefore, the key to comfortable, rapid insertion is to minimize the formation of loops and to remove them as quickly as possible when they occur. As each turn in the colon is passed, the scope tip is deflected to pull back or hook the turn, and the scope is withdrawn until the instrument shaft is straight. During withdrawal, torquing the instrument, typically in a clockwise direction, is helpful in removing loops. One way to tell that we have loops out of the instrument is to notice good one-to-one transmission. This means that when we move the scope tip back and forth with short movements, we see corresponding in-and-out movements of similar distance in the endoscopic view. This so-called one-to-one transmission is much less likely to occur when there is a bend or loop in the instrument shaft. A second way to differentiate whether a loop is present versus the instrument shaft is straight is to simply glance down and to see how much scope is in the patient. For example, we expect about 65 to 90 centimeters of scope to be in the patient when the tip is in the cecum and the instrument shaft is straight, about 40 to 60 centimeters when the instrument shaft is straight and the scope tip is in the splenic flexure, and 15 to 40 centimeters when the instrument shaft is straight and the scope tip is somewhere in the sigmoid colon. With the scope tip in the cecum in the endoscopic views and the instrument shaft straight, the scope guide will show the colonoscope tip to be in the shape of a fish hook, and you can see that only 65 centimeters of instrument are in the patient. Paradoxical motion refers to when the scope is being pushed into the patient, but the scope tip is moving out of the patient rather than inward. Even if the scope tip is not moving at all when the instrument is being pushed into the patient, we can be certain that a loop is forming. By definition, whenever we see paradoxical motion, we are forming a loop in the colon. You might as well give up whatever you are doing because it's not working. Here is another example of paradoxical motion. We can counter looping by applying abdominal pressure or by changing positions. In general, changing positions has been shown to be more effective than abdominal pressure. The most common position change is to go from the left lateral decubitus position to the supine position when there is difficulty in crossing the transverse colon or when there's difficulty passing down the ascending colon into the cecum. The most consistently reliable position change is to move from the supine position to the right lateral decubitus position when the scope tip is at the ileocecal valve and we need to move into the cecum. This maneuver can be easily accomplished by transferring the control head to the right hand and then passing the instrument shaft under the patient's legs and then reaching over the patient with the right hand to grasp the instrument shaft. Abdominal pressure can be used to resist loop formation and transmit force to move the instrument tip forward. If a loop is palpable in the abdomen, the instrument shaft should be straightened and abdominal pressure applied to the location where the loop had been previously felt. If a loop is not palpable, we typically try pressure initially in the left lower quadrant. However, it may be preferable to push up across the mid-abdomen when resisting transverse colon looping, to push on the right abdomen or flank to assist in passing the right colon, or to push up on the cecum to facilitate cecal insertion. Sometimes it's a matter of trial and error. In some cases, particularly when performing colonoscopy with a pediatric colonoscope or an upper endoscope, it may be necessary to have two assistants cover the abdomen with four hands to control looping. I almost always remove abdominal pressure when I'm withdrawing loops from the colon. I think that firm pressure on the abdomen can actually resist the unwinding of the colonoscope as the scope is withdrawn to remove loops. Let's review the fundamental principle of keeping the instrument shaft straight during insertion. We can see from scope guide that at this point there are significant bends in the instrument shaft. As we pass a turn, we pull back and remove these loops, thus straightening the instrument. Good one-to-one transmission also tells us the instrument is straight. Now we can proceed again, and as we do, new loops are forming. Again, we pull back and suction. Once we are straight, we can again move forward, and new bends form in the shaft. We'll go around a turn, and as soon as we do, we will pull back again. We'll see a lot of bends in the scope on scope guide, and these are being taken out, and we are suctioning to help shorten the colon and pleat it over the instrument shaft. Once the instrument shaft is straight, we proceed again. With the shaft now straight, our force is transmitted to the tip of the scope rather than into a loop, and this regular straightening is critical to rapid and comfortable insertion. As we proceed toward the hepatic flexure, we'll again form loops, and we will again remove them by pulling back. We should never be afraid to pull back even a long distance if necessary to straighten the scope. Only a straight scope will allow the tip to move forward efficiently. As we round the hepatic flexure and look down toward the cecum, we should resist any temptation to just push the scope forward. Rather, we must always again stop and straighten the instrument and suck air out of the colon. With the shaft straight and the lumen deflated, the scope often seems to want to fall into the cecal tip. We can see good one-to-one transmission, and now we drop into the cecal tip. Now let's put all of this together by watching a couple of insertions from beginning to end. As we start forward, we'll round the initial turns in the rectum and distal sigmoid by using the up-down control and torquing the instrument to achieve right-left turning. We'll execute a short slide-by maneuver in the up direction, which causes a bend to form in the shaft as visible on scope guide. In this case, we're going to pass one more turn, a sharp turn, which we'll execute by using the left-hand shaft grip and the right hand in the control head to turn right, and then pull back on the instrument shaft and remove the loop on the scope as evidenced by one-to-one transmission and straightening on scope guide. Again, this straightening maneuver is fundamental to rapid insertion. As we move forward again, we'll see some loop formation on scope guide, but the scope tip is moving forward, so we'll proceed for a few moments, but as soon as we pass a substantial turn, we'll stop, pull back, and suction air from the lumen. As we approach the hepatic flexure, we see a significant bend forming in the instrument shaft. We must use the left-hand shaft grip here to help negotiate the flexure, and now as we look down the right colon, we'll be smart and rather than push on the instrument, we'll pull back on it to straighten, and also we will suction air from the lumen. As we near the cecum with the loops removed, notice the fishhook appearance developing in the shape of the instrument shaft as seen on the scope guide. We're suctioning and demonstrating good one-to-one transmission, and now the scope tip is likely to fall into the cecum when we re-advance. Suction is very useful to shorten the colon and is particularly helpful when moving from the mid-transverse colon to and through the hepatic flexure and from the distal ascending colon into the cecum. Let's watch one more insertion to emphasize these principles. In this case, we're using a probe, which you see is passed down the instrument channel to provide the scope guide image rather than a special colonoscope with the scope guide sensors built into the instrument shaft. As we pass through the rectum and distal sigmoid, notice again the right-left turning achieved by the up-down control knob and torquing the instrument shaft. A bend forms promptly in the instrument, and as we pass one turn and approach the next, we pull back and straighten the instrument shaft. Notice that the mucosa is wet, so we don't need to add additional water as lubricant in this case. We're passing multiple turns in the sigmoid, keeping the shaft as straight as possible, and using up-down deflection and torquing. As bends form, we pull back and straighten as soon as we can, repeating this process over and over. We're beginning now to see some paradoxical motion as we push forward, so we know that a loop is forming, and we see that it continues to form and the tip of the scope is not moving forward. We need to get this loop out as soon as possible, so after we pass a sharp turn with the assistance of the left-hand shaft grip, we pull back to get the instrument shaft straight again. Look at the large amount of scope that's being removed from the patient from the extensive bends and loops that had formed in the scope shaft. ScopeGuide shows these bends and loops being progressively removed until the shaft is straight again. Here, you see we are pushing into a turn and forming a bend in the instrument shaft, and then pulling back on the shaft, and then you see that we can pull back also in a moment with the left-hand shaft grip while simultaneously using both control knobs to find the lumen and negotiate the turn. After the lumen is passed, we immediately pull back and suction. Our suction is slow here because the scope guide probe is in the channel, but here's a good example of how suctioning seems to pull us through the proximal transverse colon and hepatic flexure. Looking down the right colon, we again stop and suction, and I usually like to rock the scope in and out a bit here as if demonstrating good one-to-one transmission, and this often seems to shorten the right colon and bring the cecum closer to the scope tip. Notice on scope guide that the instrument shaft has all the bends and loops removed whenever possible before advancing again toward the cecum. To summarize the keys to rapid, comfortable insertion, first, keep the instrument shaft straight. Keep the mucosa wet. Use suction to shorten the colon and pull it toward the scope tip, especially in the transverse and right colons. Finally, when difficulty is encountered, move through the ancillary maneuvers rapidly until one of them works. For example, if you try left lower quadrant pressure once and it fails, immediately move to pressure in another location or rotate the patient to a new position. The fast colonoscopist knows that if an ancillary maneuver fails once, it will likely fail again. Don't hesitate to try another approach. Experts debate whether terminal ileal intubation should be part of routine colonoscopy in asymptomatic people. I would argue that it's not necessary in asymptomatic people. However, colonoscopists should practice terminal ileal intubation during routine colonoscopy until they feel comfortable with the techniques because there are times when TI intubation is mandatory. For example, all patients undergoing colonoscopy for abdominal pain and diarrhea should undergo TI intubation. The basic principles of terminal ileal intubation can be summarized as follows. First, when TI intubation is mandatory, consider a pediatric colonoscope. As in strictures and angulation, passing through a narrow opening at an angle, such as the terminal ileum, is easiest with a thin scope. The most common error that I see during the process of learning TI intubation is to not adequately deflate the lumen of the cecum and the right colon. Intubation causes the IC valve to rotate so that it points more laterally or even superiorly, greatly facilitating intubation. If the valve orifice is visible, then place the scope tip between the lips of the valve and deflect in the direction of the valve. Before deflecting, you want to actually abut against the proximal lip of the valve. The scope tip is then perfectly positioned to literally pry the lips of the valve open as the scope tip is deflected. The use of the left-hand shaft grip can again be helpful in obtaining pinpoint accuracy with tip deflection. Finally, if the valve orifice is not visible, we'll learn the bow and arrow sign to find it. With the scope tip in the cecum, the ileocecal valve orifice will be located on the same side of the cecum as the appendiceal orifice. If the valve orifice is visible, as it is on the upper left in this case, we can simply wedge the scope tip against the proximal lip of the valve and then deflect the scope tip in the direction of the valve and the scope will pop into the terminal ileum. Here's an example of how we pull back until the proximal edge of the valve comes into view and then deflect toward the valve and pop in. Sometimes the valve orifice location is indicated by a notch in the valve. In this case, it's not, but as we approach the valve with a bit of deflation, the proximal valve lip becomes visible and we can abut it and deflect toward the valve and pop in. If the location of the valve orifice is not already obvious, then the bow and arrow sign can be used. The bow and arrow sign makes use of the appendiceal orifice. The appendiceal orifice, with its crescent shape, forms the bow, and an imaginary arrow is drawn through the middle of the bow. In the great majority of cases, this arrow will point in the direction of the ileocecal valve orifice. With the lumen deflated, and the scope gradually withdrawn and deflected in the direction designated by the arrow, the proximal lip of the valve will come into view, and the scope tip can then be deflected into the terminal ileum. Here's another example of the bow and arrow technique. The appendiceal orifice is the bow, and an arrow is drawn through the middle of the bow. Now we deflate and pull back slowly until the proximal lip of the valve orifice is visible. There it is now. With that lip abutting the scope tip, we deflect in the direction of the valve. If we see the villi of the TI, we know we're going in the right direction, and if we pull back on the scope a tad while continuing to deflect toward the valve, we pop in as we have here. The choice of sedation can impact the technical performance of colonoscopy. There's no aspect of colonoscopy for which there's more cultural variation between countries than in the use of sedation. For example, in France, colonoscopies are routinely performed with propofol sedation delivered by anesthesiologists. In Scandinavian countries, such as Finland, more than 90% of colonoscopy is performed without sedation. In the United States, only about 1% of colonoscopy is performed without sedation. In my unit, about 2% of patients ask for colonoscopy unsedated, but if we inform patients that unsedated colonoscopy is available, 7% request an attempt at colonoscopy without sedation. Most colonoscopy in the United States is still performed with moderate sedation, formerly called conscious sedation, though this is changing rapidly with the growth in anesthesiologists administering propofol sedation. Moderate sedation is usually achieved with combinations of opioids and benzodiazepines. It's quite possible to perform colonoscopy with moderate sedation using propofol if propofol is given in combination with low doses of narcotics and benzodiazepines. Propofol as a single agent cannot be titrated to moderate sedation for colonoscopy because patients will exhibit reflex withdrawal responses to pain. However, low doses of opioids and benzodiazepines will block these pain responses and allow propofol to be given in much smaller doses, given less frequently, and still titrated to moderate sedation with no significant loss of patient satisfaction compared to propofol alone and with recovery times that are actually faster compared to when propofol is given as a single agent and titrated to deep sedation. About 25% of colonoscopies in the United States are now performed with propofol-based sedation. More than 90% of propofol administration is by anesthesia specialists alone. Opioids and benzodiazepines are also frequently titrated to deep sedation by endoscopists. There is probably more risk associated with opioids and benzodiazepines in deep sedation than is associated with propofol. In general, opioids and benzodiazepines are optimally targeted to moderate sedation. Propofol has been administered safely by gastroenterology groups within and outside the United States. This practice, sometimes called endoscopist-directed propofol, typically involves the administration of propofol titrated carefully by a trained registered nurse. In Germany and Switzerland, this has become the most commonly used form of sedation for colonoscopy. In the United States, the growth of endoscopist-directed propofol has been impeded by several factors, including the propofol package insert in which the FDA states that propofol should only be administered by persons trained in general anesthesia, and a follow-up statement reinforcing this position by the FDA in 2010. In addition, regulations issued by the Center for Medicare and Medicaid Services in December 2009 and updated in January 2011 make the administration of deep sedation the sole domain of anesthesia specialists. Although propofol can be titrated to moderate sedation when given in combination with opioids and benzodiazepines, consistent positions taken by the American Society of Anesthesiology stating that propofol use always implies deep sedation have made the creation of institutional policies endorsing propofol for moderate sedation quite difficult. For unclear reasons, CMS never instituted regulations for the use of propofol in ambulatory surgery centers. Certain patient factors predict an increased likelihood of successful unsedated colonoscopy. These include male gender, since on average colonoscopy is easier to perform in men than women, older age, the absence of pre-existing abdominal pain, higher educational level, and lower pre-procedure anxiety level, which is asterisked here because we typically don't have a formal measurement of that prior to starting the procedure. In my opinion, unsedated colonoscopy should be patient-motivated. Patients should not be encouraged or pressured to undergo unsedated colonoscopy. The widespread use of sedation in the United States enhances the public reputation of colonoscopy as being comfortable. One study showed that patients who underwent unsedated flexible sigmoidoscopy were more than twice as likely as patients in the same medical practice that had undergone sedated colonoscopy to say that they would not return for another screening examination. However, the advent of water immersion colonoscopy has made it more reasonable to let patients know that unsedated colonoscopy is a viable option. Several factors can facilitate unsedated colonoscopy. First and foremost is the use of water immersion. This technique will be reviewed in detail in a later section of the DVD. With or without water immersion, unsedated colonoscopy may be facilitated by the use of a pediatric colonoscope, especially in women, because increased flexibility of the scope means less force is exerted on the mesentery, and therefore there is less discomfort. When the procedure begins, warn the patient about sensations they are likely to experience, including the urge to defecate, unusual abdominal sensations, and pain. They should be reassured that loop formation will be kept to very short intervals, and use the right-left control rather than torque to turn during insertion in unsedated colonoscopy. In moderate sedation, the patient can respond to verbal commands. They can assist with position change. They are more likely to protect their airway in the case of regurgitation than are patients in deep sedation, and they tend to maintain anal sphincter pressure and thus retain air better during withdrawal. Deep sedation has the advantage of being able to push through loops for which a patient in moderate sedation would sometimes not be able to tolerate the associated pressure. There are a few patients who cannot tolerate colonoscopy in moderate sedation, which is an additional advantage for deep sedation. Disadvantages of deep sedation are that when turning is required, the patient must have their position changed by the medical assistants and the medical team. The risk of aspiration is higher during deep sedation, particularly when the patient is in the supine or right lateral decubitus position, and strict attention must be paid to airway protection. In addition, during withdrawal, patients do not retain insufflated air as well, and it may be necessary for the assistant to mechanically hold the buttocks together so the patient can retain insufflated air or carbon dioxide. Although pushing through loops is more feasible in deep sedation, it's generally not advocated to do that as a routine practice. As we've discussed, loop reduction is critical to safe, comfortable, and rapid insertion of the colonoscope. However, there are times when loop formation is the only way to advance the scope, and this may be better tolerated with deep sedation by some patients. Deep sedation is not the optimal sedation for training in colonoscopy since it may encourage pushing through loops. Some people feel that it is important to have the patient's verbal signal of pain to warn the endoscopist about impending perforation. I believe that that is not true, and rather, the endoscopist must feel with his or her hand when there is fixed resistance to colonoscope insertion and stop the insertion at that time, regardless of the patient's level of sedation. An emerging method of performing insertion that should be learned by all colonoscopists is the water immersion technique. The basic idea is very simple. Rather than insufflating the colon with gas during insertion, no gas is insufflated into the colon. In fact, we remove any gas that we encounter, we are going to distend the lumen with water only, and this water distension will allow us to pick up the visual cues and perform the insertion. Now, why should we know about the water insertion method? First, there is clear evidence from randomized controlled trials that if patients are undergoing colonoscopy without sedation, or with minimal sedation, that they experience less pain with the water immersion method than with the traditional gas insufflation method. Secondly, water immersion can greatly facilitate sacral insertion, and in my own opinion, the instance in which it does this the most is the patient with a redundant colon, and I'll come back to the reasons why it's helpful under those circumstances. In the literature, you'll also see the term water exchange, which is really just another form of water immersion. So we're demonstrating here the technique of water immersion. The first step is to go to the gas source, whether you're using air or carbon dioxide, and turn that off. In true water immersion, we're going to use water only and no gas of any kind. In fact, when we encounter gas during the insertion, we're going to suction that out. We just passed a large polyp at the rectosignoid junction. Now, I prefer to put the water in using the water jet. So put the scope into the rectum and just start pumping water, filling the lumen enough that you can see the luminal direction using the same visual cues that you would use if you were using gas. Now, this works best when the prep is good. If the prep is poor, the fluid is going to have a dark green color, and then you have to both pump water in and suction simultaneously, and that's been referred to as water exchange, but the process is basically the same. Now, some practicing colonoscopists use a variation of water immersion technique where they basically have the gas on and they use a combination of water and gas, and that actually, in my experience, can have value and merit because the water still lubricates the colon so that the hydrophilic forces between the scope and the mucosa are better. It's easier to push the scope in because it just slides easier. But also, if you're going in with gas insufflation and you encounter a difficult turn in the sigmoid and your angulation is quite severe, you can simply start pumping water in and sort of change the dynamics between the scope and that angle, lubricate the colon better, and you'll often find that that will help you to pass a specific turn. So we've sort of got all variations of the use of water, but I think one of the keys is that basically the use of water is oftentimes helpful. Now, most of the benefits of water immersion occur in the left colon, so one of the issues is when do you stop, and I think in many instances you can go all the way to the cecum because you'll be able to identify underwater the ileocecal valve or the appendiceal orifice, or you'll just reach an area where you can't find the lumen anymore, and you may be in the cecum there, or you may be in the transverse colon. And especially if the prep is poor and the transverse colon luminal diameter is big, it can be better at that point to stop and convert over to gas because, as I said, most of the insertion benefits have been realized at that point. Now, when should water immersion from the get-go be considered? In my own practice, I do it routinely now when I'm performing unsedated colonoscopy. We do have a small percentage of patients who request the examination with no sedation, and there is where our randomized controlled trials really show the greatest benefit because you get a reduction in pain, you get a reduction in the need for salvaging the discomfort by providing medications, or if you're using very low levels of sedation from the get-go, there's a reduced need to add additional medications. But the other instance in which it's very helpful is the patient where you know the colon is very redundant and elongated, and we have quite a few referrals for this. In the past, we've had to use some sort of special ancillary technique like an endoscope or an overtube in about a third of the cases, but we've recently reported in GI endoscopy that we could reduce the need for that to about 5% by the use of water immersion. So we're filling the colon. We're just now getting to the cecum. You can see the ileocecal valve there appearing on the upper left, perhaps a little bit worse here. We could stop at this point and begin to insufflate gas because the withdrawal phase of the examination is still going to be done with gas insufflation. So the basic technique is we use water filling to get to the cecum or to get through the left colon, and then once we get up to the cecum, then at that point, we typically are going to stop and begin the process of suctioning fluid. Now, surprisingly, although you may have put a large volume of fluid in, you usually find that there's not quite as much as you would expect to suction on the way out. So you can see at this point, you can start to see some gas being insufflated in the colon. In this particular case, the water actually is probably helping to improve the prep as it often does, and that's kind of a side benefit at times with water immersion. I don't think that aspect of it is really worth it. We really should have a good enough prep up front that we can get a good examination with air insufflation. But here you can see we're now filling the cecum with gas, and now we're going to do our usual withdrawal technique to accomplish an effective examination. This brings us back to the question of why water immersion is helpful in redundant colons. Most colonoscopists have had the experience of performing a colonoscopy in which you get to the hepatic flexure or the mid-ascending colon, or perhaps even the level of the ileocecal valve, and you're creating a loop in the colon, and you're simply out of scope, and you can't go forward. You really don't have additional scope to put into the patient. When we pass the scope through a straight and sigmoid colon that is filled with water, we use up less scope length compared to what we have with gas insufflation. If the colon is both elongated and dilated, which it is sometimes, sometimes the actual luminal caliber is increased. With air, when the colon is extended, we actually use more scope going around each curve than we do when we fill it with water, and we keep the luminal diameter narrow and straight. And so we're simply less likely to run out of scope in the right colon. I mentioned earlier in the informed consent section that perforation is often considered the most serious complication of colonoscopy. Most recent data from the Medicare database indicates that the perforation rate overall for colonoscopy in the Medicare population is still about 1 in 500, and the risk of perforation increases with each five-year interval of advancing age, and also increases with diverticular disease. There are two basic mechanisms of perforation during insertion. The first of these is mechanical rupture by the colonoscope, which usually occurs when the side of the colonoscope ruptures the colon in the vicinity of the rectosigmoid junction. The most important step in avoiding this type of perforation is to not continue to push the instrument when fixed resistance is palpated by the hand on the insertion tube. Fixed resistance is the sensation that the colon will not stretch further as pressure is applied. No matter whether the lumen is in view or not, you must stop when you feel fixed resistance. You must also stop when you feel fixed resistance, whether the patient is alert or in general anesthesia, and whether they report pain or not. Options to overcome fixed resistance include withdrawing the instrument and completely straightening the colon and then reattempting the insertion, or changing to a thinner, more flexible instrument. Filling the colon with water rather than gas may also be of value. Mechanical perforations may also occur when the colonoscope tip is forcefully pushed against a diverticulum. Remember that not all diverticula are small, and occasionally diverticula will mimic the lumen even for an experienced examiner. Mechanical perforations occur more rarely when the colonoscope tip is pushed against a stricture that is too tight to allow passage of the scope. Continued pushing can cause the tip to forcefully slip off the side of the stricture and dissect the normal colonic wall. This type of perforation can be avoided by either first balloon dilating the stricture, or by passing a guide wire through the stricture before attempting passage, as shown in the animation. The guide wire will prevent the scope tip from slipping off the side of the stricture. The second major category of perforation that can occur during insertion is barotrauma perforation. Barotrauma perforations occur from air pressure. The first step in preventing barotrauma perforation is to make sure that the air insufflation control on the colonoscope device is set so that it will not continue to pump air at pressures that exceed the sequel bursting pressure. Secondly, take it easy on the air during insertion. Enough air should be insufflated so that the lumen can be clearly identified, but additional air insufflation will lengthen the colon and make insertion more difficult, as well as increasing some risk and discomfort for the patient. Be particularly careful about air insufflation after passing a tight stricture or angulation in a patient with severe diverticular disease. In such a patient, air may not be able to escape around the colonoscope in order to exit through the anus. If the ileocecal valve is competent to air, a closed system can develop. Barotrauma perforations can occur after the colonoscopy has been completed. The usual scenario is the patient with severe diverticular disease and a large amount of air trapped proximal to significant sigmoid angulation. Such a patient can have prolonged colonic distention and eventually develop sequel ischemia or perforate spontaneously in the same manner as the patient with Ogilvy syndrome. At the end of the examination, always check the abdomen for the level of distention, especially if you're using air rather than carbon dioxide. If the abdomen's tight, consider re-intubating the proximal colon and removing air, especially in a patient with a sigmoid stricture or complex diverticular disease who may have difficulty passing that air. Do not let a patient who's severely distended or uncomfortable from air and unable to pass gas leave your recovery room without decompression. Essentially, all of the problems that are related to gas distention during colonoscopy can be avoided or nearly entirely avoided by the use of carbon dioxide insufflation rather than air. And we now use carbon dioxide for all of our colonoscopy cases. Carbon dioxide is absorbed from the colon about 150 times faster than air. The advantages of carbon dioxide include the following. One, patients are much more comfortable in the recovery area. They are discharged faster. And this is especially true for doctors who like to insufflate the colon with plenty of gas during withdrawal. And I think that is overall the right thing to do. I see some people performing colonoscopy who are very focused on removing gas as they are coming back through the colon. This can be a less than optimal maneuver, because to see well, you have to put gas in. And if you remove it from one section of the colon and then inflate the next section, of course, depending on the patient's position, a fair amount of it is going to move into the more proximal colon. You don't have to worry about distention with carbon dioxide. Two, you do not have confusion in the recovery area about whether there's been a complication. For example, if you take out a very large polyp, it's very worrisome if the patient has a lot of discomfort in the recovery area, because they could already have developed peritoneal irritation or even perforation. So if a patient has pain after using air, a lot of people are thinking, oh, this is probably from air distention. Whereas if the patient has a lot of pain after using CO2, probably something bad is going on, and they need to be worked up. Recent data from Australia indicate that patients who have had large colorectal polyps resected are less likely to be admitted to the hospital for abdominal pain without perforation when CO2 is used compared to when air is used. Third, there's probably a decreased risk of complications with carbon dioxide insufflation. These include the following. Reduction in the chance of barotrauma perforations, as already discussed. Most common in the patient with sigmoid colon diverticular disease that's quite severe and which prevents patients from passing gas after the procedure. These patients can go home, eat additional food and drink, and then if they have a competent ileocecal valve and they fill their cecum, they may get increased pressure and develop a delayed perforation. In the case of performing a complex procedure such as stent placement or decompression of Ogilvy's syndrome, it's better to take gas out and not put additional air in. In my experience, you get a better decompression of Ogilvy's syndrome when using CO2. For a difficult stent placement, it's just wiser to not pump in a lot of air that could go proximal to the obstruction and create a lot of distention. So CO2 is a good idea in that circumstance. When you're removing a large sessile polyp, the use of CO2 can reduce the tension on the wall or the stretch on the polypectomy defect. If you see an area in the defect that looks like muscle injury, you really want to be careful about insufflation and CO2 is safer in that situation. A good colonoscopist must have both a high secal intubation rate and a low perforation rate. Following the rules we've discussed can greatly improve the safety of colonoscope insertion. The next aspect of colonoscopy to discuss is withdrawal technique. Withdrawal is the most important phase of colonoscopy since the purpose of colonoscopy is usually to search for disease and withdrawal is when we search. In the section on informed consent, I discussed recent evidence indicating that colonoscopy is not generally as effective against the development of colorectal cancer as we thought, particularly in the right colon. There are several a priori possibilities for why a patient could undergo colonoscopy with apparent clearing of the colon of polyps and then present within the next few years with colorectal cancer. One of these is biologic variation in the growth rates of tumors. We might expect, for example, that poorly differentiated tumors could grow quite rapidly and we're quite certain that microsatellite unstable tumors can develop from small adenomas in a short period of time. There's evidence that cancers developing after colonoscopy are more likely to be in the proximal colon and have an increased prevalence of microsatellite instability and SEMP high. A second potential explanation is technical limitations of colonoscopy. Older studies with barium enema and recent studies with CT colonography demonstrate that polyps on the proximal sides of folds and flexures are more likely to be missed during colonoscopy. Indeed, there may be some lesions that are so strategically positioned that even the most careful colonoscopist would miss them. In addition, there are lesions that are so flat or subtle that they may be invisible in white light to some or all endoscopists. A third explanation is that polyps were detected but incompletely removed. As will be discussed in an accompanying DVD on polypectomy, it is critical to perform close follow-up of larger sessile polyps that are removed in piecemeal fashion. Another potential cause of missed lesions is ineffective bowel preparation. In data from the Clinical Outcomes Research Initiative project sponsored by the ASGE, about 25% of colonoscopies performed in the United States were considered to have inadequate preparation. The optimal combination of safety, effectiveness, and tolerability in a bowel preparation has yet to be determined. However, my anecdotal impression is that many colonoscopists accept suboptimal preparation as a routine. Ideally, the colon should be cleaned of solid stool and mucus from the appendiceal orifice to the dentate line. This slide depicts an excellent colon preparation which we should strive for on a routine basis. Many endoscopists accept the presence of tenacious chyme and mucus in the cecum ascending colon. However, this material is problematic because a disproportionate percentage of flat neoplasms are located in the proximal colon. And flat neoplasms are very common even in the United States and Western countries. The key to clearing chyme and mucus from the colon is to give part of the preparation on the day of the examination. For any bowel preparation, part of it can be given a few hours before the time of the scheduled colonoscopy. This will clear the right colon and cecum of chyme and mucus that have passed out of the small intestine and accumulated there following the initial dose of preparation. The American Society of Anesthesiology allows clear liquids to be taken up till two hours prior to the time of sedation. Thus, a requirement to keep patients NPO after midnight the night before the examination is misplaced. To summarize available data, poor preparation has been associated with more missed small and large polyps, longer times to the cecum, and longer times for withdrawal, and more costs because patients are often required to repeat their examinations at earlier intervals than are recommended in current screening and surveillance guidelines. An important key to excellent bowel preparation is for some of the prep to be ingested on the day of the examination. The final element that may cause patients to return with cancer after colonoscopy is inadequate application of colonoscope technology by some examiners. The evidence that some examiners use suboptimal examination techniques includes both variable detection of cancer and variable detection of adenomas, as was shown on the slide from earlier in the DVD. And we now also have evidence of variable detection of serrated lesions. This table shows that members of the same gastroenterology group, not uncommonly, have three to six-fold differences in their adenoma detection rates. If adenoma detection is expressed as adenomas per colonoscopy, the difference between examiners becomes even greater. This slide shows this effect for nine endoscopists at Indiana University Hospital. Between the top detecting endoscopist A and the lowest detecting endoscopist B, there's a four-fold difference in adenomas per colonoscopy. Endoscopist A has a higher adenoma detection rate, but also detects more patients who have two or more adenomas and more patients who have three or more adenomas, et cetera, magnifying the differences in detection identified by the adenoma detection rate. Therefore, the adenoma detection rate does not fully grasp the variability between examiners and the extent to which low-detection endoscopists miss lesions. Two recent studies have examined variability in detection of serrated lesions between members of the same group and show even greater differences between endoscopists in proximal colon serrated lesions detected than was seen for adenoma detection throughout the entire colon. This suggests that low-level detectors miss an even higher percentage of serrated lesions than they do conventional adenomas. There is limited evidence about what constitutes optimal withdrawal technique. In an Indiana tandem colonoscopy study, there were 26 colonoscopists who participated in the examinations. The lowest and highest measured miss rates occurred in two experienced colonoscopists who had each performed more than 10,000 prior colonoscopies. Subsequently, 10 consecutive withdrawal examinations by each of these endoscopists were videotaped, and the videotapes were shown in a random order and in a blinded fashion to four independent gastroenterologists who ranked segments of the colon for the quality of examination of the proximal sides of folds, adequacy of distention, cleaning up of dirty areas, and taking adequate time to examine. The colonoscopist with the lowest miss rate had highly, significantly superior quality scores for each of the four criteria and by each of the four independent GI colonoscopists. Thus, better detection is associated with careful examination of the proximal sides of folds and flexures, adequate distention, cleaning up dirty areas, and taking adequate time to examine. These principles of high-quality technique have been demonstrated in several subsequent studies of video recording colonoscopy examination technique. Based on this evidence, the U.S. Multi-Society Task Force on Colorectal Cancer and a combined task force from the ASGE and ACG have made recommendations regarding detection during colonoscopy. Individual endoscopists should have their prevalence rates of adenomas measured in patients undergoing colonoscopy who are 50 years of age and older and one or more adenomas should be detected in at least 25% of men and 15% of women. In addition, and particularly in those individuals in whom detection rates of adenomas are below standard, the withdrawal phase of colonoscopy should average at least six minutes in normal colonoscopies in which neither biopsy nor polypectomy are performed. During withdrawal, the instrument shaft is typically without loops and a straight instrument allows excellent control of tip deflection, which we can typically achieve using the up-down control and torque on the shaft to achieve right-left turning. The principal work of withdrawal is taking time to explore the proximal side of each haustral fold for which we must have superb control of tip deflection. High-level detection of neoplasia requires excellent bowel preparation as demonstrated here. Clearing of mucus and chyme from the right colon allows us to see even the most subtle flat lesions. Good withdrawal technique has an obsessive-compulsive character to it. The endoscopist develops a three-dimensional sense or map of the shape of the colonic lumen as it's created by haustral folds and turns and valves. The endoscopist seeks to expose the mucosa hidden by these obstacles. The mucosa in the inner haustral valleys and on the proximal sides of haustral folds must be explored by deflecting the tip into the inner haustral valley and sweeping right and left and pulling back to flatten the haustral fold as needed. As the scope is withdrawn and each haustral fold is identified, the endoscopist pushes the scope forward again and deflects to see the proximal side of the fold. This process is done systematically with the endoscopist utilizing the sense of 3D shape of the colon to be certain that each fold and side of the colon is inspected carefully. Pools of fluid are suctioned to allow inspection of hidden mucosa. The colon must be adequately distended with air or carbon dioxide so that mucosa is not hidden by collapse of the colon wall. In teaching fellows to perform careful withdrawal, I often tell them to, quote, work the folds, close quote. Again, this refers to the process of going back in as each haustral fold is passed to explore the proximal side of that fold. Even the most careful colonoscopist will miss small adenomas with current technology, but we often use the expression aim small, miss small. This term is sometimes used in learning to shoot a gun or pitch a baseball, where it means to keep the target you're aiming at small so that if you miss, you'll miss by very little and still be effective. In colonoscopy, it means that if we work to detect even the tiniest lesions, we'll be unlikely to miss large lesions that everybody would agree are clinically important. This process of careful examination, though not very interesting to carry out, is fundamental to our responsibility to our patients to provide the best examination possible. Because of the lower protection of colonoscopy against right-sided compared to left-sided cancer, there's a lot of discussion about whether or not we should be doing something extra to try and detect lesions on the right side. Now, obviously something extra includes having a very high-quality bowel preparation and includes a colonoscopist who's trained to recognize flat and depressed adenomas and also serrated lesions, all of which are more common in the proximal colon. Special maneuvers won't be helpful if doctors can't recognize the lesions that are there. What are some of the maneuvers that can be used? Well, one of the things that's often discussed is retroflexion in the right colon, and it's become relatively common for some people to do it on a routine basis. So we're gonna talk a little bit now about the performance of retroflexion in the ascending colon. This is usually done in terms of the examination as part of a second examination. That is, we've already conducted a first careful examination from the appendiceal orifice back to the hepatic flexure, and then we're considering doing a second examination. And oftentimes, we might choose to do that in retroflexion. However, there's a randomized controlled trial that shows that performance of that second examination in the forward view is actually just as effective as doing it in retroflexion. Retroflexion is kind of attractive because of the theoretical reasons of being able to see better on the proximal sides of folds. It's certainly a very important skill to have for the performance of complex polypectomy, but we don't know that it's really better than doing a second examination in the forward view. Now, when should we consider doing a second examination at all? I think the best reason is when we've already found things on the first examination. In this case, I removed already a large serrated lesion from the hepatic flexure, and there's a second serrated lesion, which is present right here, sort of in the distal ascending colon. So the more we find on the first examination, the greater the rationale for performing two examinations. Now, what about the choice of doing the second examination in retroflexion as opposed to doing it in the forward examination? Well, there are certain predictors that we're gonna have a relatively easy time getting into retroflexion. The first one is that we've got good one-to-one transmission, which means that we don't have a loop in the instrument. The single strongest predictor that we're gonna have a hard time performing retroflexion in the ascending colon is if we have a loop in the instrument. So in this case, we've got good one-to-one transmission. So now we'll go ahead and do our second examination, and in this case, I'm gonna do it in retroflexion. So what's the technique? Well, the first thing is we wanna move down close to the ceca, find a space that's relatively open, and then we're gonna go max in two directions, in this case, max up and max left, and then put our hand under the scope and roll over, and that will bring us into retroflexion. Now we're looking up the ascending colon, and then we just come back and do a careful examination of the proximal sides of folds. Here is our large serrated lesion that we have yet to remove, and as we move up toward the hepatic flexure, we can see a polypectomy site from where we've already removed a lesion and clipped it. So somewhere up close to the hepatic flexure, when we find a relatively open space, we will have completed our retroflexed exam, and we wanna come out of retroflexion. Now, I recommend that to do this, we release the controls, and then rather than just let the bending section push up against the wall, which would put some pressure against the wall, we're gonna release the controls and also pull back on the insertion tube so that we unwind like this with this motion, thereby putting less pressure on the wall. So here we are in retroflexion, I'm releasing the controls, and notice I'm pulling back on the insertion tube, letting things straighten out a little bit so we don't put too much pressure on, and then we're back in the forward view, and we're done with our retroflex second check of the colon. Now I'm gonna go ahead and proceed with removal of the second serrated lesion. As a safety measure, I would say that you should not, with a standard colonoscope, either adult or pediatric, pull a scope in retroflexion all the way out through the colon, because there's some chance it's going to get stuck in the left colon. You may be able to pull it all the way back to the splenic flexure, but once you get into the left colon, there are reports of perforating the colon during unwinding of the scope. If you need to retroflex for removal of a polyp in the left colon, use an upper endoscope, because it's much narrower and has a much tighter turning radius. We'll continue this process of careful examination now through the whole colon. Here are a few comments in real time about examination of the rectum. Now as we come back in the rectum, we're gonna continue the same basic principles of withdrawal, but we have to remember these rectal valves can really be quite prominent. They can hide quite a bit of mucosa. So I like to examine the rectum first in the forward view, looking behind each one of these rectal valves. And then once we've done that, we come all the way back to the anus. We're here at the dentate line. You can see that in most cases, we actually can see in the forward view the entire rectum. So it's quite common for people to perform retroflexion. So that maneuver's done by simply inserting the scope tip up into the mid rectum. And then we wanna max the scope in the up direction. And I usually go in the left direction, but in two directions. Then put our hand under the instrument shaft, rotate over, and advance. And we get a good retroflexed view. You can see some internal hemorrhoids. And then I would take this photograph for documentation purposes. And then the exam's complete. Remove some air for patient comfort, and we're finished. This slide summarizes current knowledge regarding optimal withdrawal technique and time. Colonoscopic imaging technology is in flux at this time. As awareness has grown that colonoscopy can miss significant lesions, new technologies have been developed or are under development or have even been commercialized. Some of these show promise for improving detection or reducing the wide variability between endoscopists in detection. None of them has yet met the test of being both significantly effective and practical to apply. Until that happens, and likely for a good while, there will be no substitute for applying careful standard withdrawal technique and measuring detection rates and withdrawal times in continuous quality improvement programs. The final issue that we will discuss is the approach to the colon in which sacral intubation must be obtained but proves difficult. The approach I will summarize here reflects my experience in performing colonoscopy in patients who are referred after failed attempts at colonoscopy by other gastroenterologists or surgeons. This experience has appeared in several sequential publications. Very difficult colons can, for the most part, be grouped into two categories, the very difficult sigmoid and the very redundant or elongated colon. The difficult sigmoid colon is typically one in which there is very severe diverticular disease, often associated with marked angulation that is difficult to pass. The first step is to attempt passage with a pediatric colonoscope. If that fails, the next step is to try a thin upper endoscope. Thin upper endoscopes can almost always be passed through angulated areas in the sigmoid. Once the angulation is passed, a vigorous attempt should be made to pass the upper endoscope to the cecum, which should be successful in about two-thirds of cases. Because of the flexibility of this endoscope, aggressive abdominal pressure must typically be applied, sometimes with four hands, that is, two assistants applying pressure in order to cover the abdomen. If the cecum cannot be reached with a thin upper endoscope, the next step is to exchange the thin upper endoscope for a pediatric colonoscope over a guide wire. This technique of guide wire exchange should not be used by novices in colonoscopy. With the upper endoscope as far in the colon as can be achieved, a very long wire is passed into the colon, and the upper endoscope is withdrawn. The wire should have a soft tip, and it must be stiff so that it will not coil during withdrawal of the upper endoscope. I prefer a colon-length Savory wire for this purpose. After the upper endoscope is withdrawn from the patient, we have a stiff guide wire with the tip in the transverse or right colon, with no loops in the wire, and with the wire exiting the patient's rectum. The wire must now be fed backwards through a pediatric colonoscope. The Savory wire has a stiff tip, and if it's fed backwards directly through the scope, the stiff tip will impact in the angle in the control head. In order to get the wire back through the pediatric colonoscope, a snare is passed anti-grade down the scope until it exits from the scope tip. The wire is then fed into the snare sheath, and then pushed backwards up the scope channel. The snare sheath protects the hard tip of the wire and guides it through the bend in the control head. We are now ready to advance the pediatric colonoscope over the guide wire. The scope will follow the guide wire nicely, but it is perfectly reasonable to use tip deflection to help pass over folds and around angles. How the technique works to allow passage of the pediatric colonoscope is uncertain. It may be that the wire itself straightens the angle in the colon. As the scope is advanced over the wire, it is often helpful to withdraw the wire in small increments equal to the amount of forward advancement of the colonoscope through the colon. Once the pediatric colonoscope is well beyond the angulation, the wire can be fully withdrawn, and the pediatric colonoscope is advanced to the cecum using standard techniques. In patients with a redundant colon, water immersion greatly facilitates insertion. By preventing the use of the full length of the insertion tube, by keeping the colon shorter, particularly the left colon, the first attempt is typically made with an adult colonoscope and good standard technique with careful attention to removing loops as each corner is passed. My preference in this instance is for the variable stiffness colonoscope despite the lack of proof of its effectiveness. If this attempt fails, including the use of water immersion, I typically use an external colonoscope straightener, such as the Olympus 60 cm overtube shown at the bottom of this picture. If the cecum still can't be reached, the next step is to try an enteroscope with or without an enteroscope straightener. The enteroscope straightener is shown at the top of this picture. Finally, if that fails, the single or double balloon enteroscope can be tried. Although I don't personally use balloon-assisted enteroscopy, this approach may replace the use of colonoscopes with straighteners with additional experience. That concludes our comments about the technical performance of colonoscopy. I invite you to review an accompanying DVD sponsored by the ASGE on polypectomy. Best wishes to you and your patients for safe and effective colonoscopy.

colonoscopy

informed consent

colonoscope

tip deflection

anatomic landmarks

sedation

withdrawal technique

adenoma detection rate

detection rates

cancer prevention

terminal ileum

risks and complications

missed lesions

scope insertion