Colonoscopy Quality and Impact on Cancer Risk: My Experience at Holy Cross

Colonoscopy Quality and Impact on Cancer Risk: My Experience at Holy Cross

  • Posted Mar 08, 2019
  • hchadmin

By Dr. Patrick Amar

Colon cancer is the third leading cause of cancer death and is preventable in up to 95% of cases. The purpose of a screening colonoscopy is to prevent the occurrence of colon cancer in the future. Without screening of any sort, the likelihood of a patient at average risk of getting colon cancer is 1 in 18. Colonoscopy is a tremendously powerful cancer prevention tool but unfortunately suffers from a number of potential pitfalls. It is both appropriate and critically important for patients and referring providers to inquire about quality measures when selecting a physician to perform colonoscopy.

I have collected data on over 1,500 consecutive colonoscopies which I personally performed over a two-year period. For each of these procedures, I collected data on which portion of the colon was reached, the quality of the bowel preparation, the number of polyps removed and the recommended interval for repeat colonoscopy. For any patient with polyps removed, I subsequently correlated these numbers with their pathology results and revised the polyp counts accordingly, to reflect only those polyps which were found to be adenomatous or serrated. These two particular pathology findings are directly correlated with cancer potential and are the target lesions for which we ask patients to undergo colonoscopy.

The first and most obvious measure in assessing quality of colonoscopy is the rate of completion of the exam. Clearly, examining only a portion of the colon leaves patients at risk for precancerous polyps in the areas not visualized. A colonoscopy is considered complete when the cecum (beginning of the colon) is intubated, as identified by several landmarks. The benchmark minimum recommended standard for this measure is 95% cecal intubation. My cecal intubation rate is 99.6%.

Cecal intubation is not a perfect measure of a complete colonoscopy, however. The cecum can sometimes be mistaken for other parts of the colon, especially the hepatic flexure. This leaves the potential for entire large segments of the colon not to be examined at the time of a colonoscopy. The solution that many gastroenterologists have proposed, to ensure 100% certainty of a complete colonoscopy, is intubation of the end of the small intestine (terminal ileum), which has a distinctly different appearance than the colon on endoscopic evaluation. This also affords the opportunity to evaluate for small bowel pathology which may otherwise be undiscovered (I have diagnosed several cases of Crohn's Disease in this manner). There is no recommended benchmark for intubation of the terminal ileum on screening colonoscopies. Nonetheless, I can proudly say that I have a 98% rate of intubating the terminal ileum for screening colonoscopies, thus dramatically reducing the likelihood of an incomplete examination.

The next and perhaps most important measure in determining quality of colonoscopy over a large number of procedures is the likelihood of discovering precancerous polyps. As noted above, these are either adenomatous or serrated on pathology. Patients with numerous or larger polyps are at an especially increased risk of colon cancer. The measure most commonly used is referred to as Adenoma Detection Rate (ADR), or the percentage of patients who are found to have at least 1 precancerous polyp at the time of screening colonoscopy. The minimum benchmark recommended for adenoma detection rates is 25% among male patients and 15% among females. The national average for ADR is 26% among community physicians and 29% among physicians in academic settings. My personal adenoma detection rate (ADR) is 54%.

The reason that this number in particular matters greatly is that there is tremendous variability in adenoma detection rates among different physicians. A recent study published in the New England Journal of Medicine which examined over 300,000 screening colonoscopies revealed Adenoma Detection Rates ranging from 7.4 to 52.5%. Analysis of this data revealed a dramatic decrease in the likelihood of patients being diagnosed with cancer up to 10 years after their colonoscopy was performed ("interval cancers"), if the performing physician had a higher ADR.  For every 1% increase in adenoma detection rate, patients had a 3% decreased risk of cancer over the following 10 years. Patients undergoing screening colonoscopy by the highest-performing physicians by ADR rates had up to a 69% incremental decrease in "interval cancer" risk compared to the average. This is over and above the baseline reduction in colon cancer risk offered by an "average" colonoscopy.

This measure, however, also has its limitations. The gastroenterology literature discusses the danger of a "one-and-done" colonoscopy. This refers to the tendency for some physicians to overlook adenomas beyond the first one discovered, either due to inattention or in the interest of moving a colonoscopy along. This gives rise to a measure called "adenomas per positive," which refers to the total number of adenomas discovered in patients who had at least one precancerous polyp. This is meant to counteract the tendency for a "one-and-done" colonoscopy and ensure that patients are getting the best cancer protection possible. The published data reveals a national average of 1.9 Adenomas Per Positive among academic centers and 1.65 among community physicians. My personal Adenoma Per Positive rate is 2.44, meaning that in patients who have at least 1 precancerous polyp, the likelihood is that I will find between 1 to 2 additional precancerous lesions.

There are polyp types which do not present an increased risk of cancer to patients. The most common type is a hyperplastic polyp which carries no precancerous potential unless it is part of a specific and rare polyposis syndrome. Likewise, the finding of a "mucosal excrescence" or other benign pathology findings do not represent precancerous tissue and are not counted in a physician's adenoma detection rate. These polyps most frequently occur in the rectosigmoid colon and are often removed by physicians, inadvertently leading many patients to think that they may beat increased risk of colon cancer. Removal of these polyps does not benefit patients in any way and does place them at some increased risk of post-procedure bleeding.

There are a number of key factors which have been found to impact a physician's adenoma detection rate. These include appropriate withdrawal time of the colonoscope during examination of the colon, the quality of the patient's bowel preparation, irrigation and suctioning of any stool residue at the time of examination and appropriate use of available advanced technologies to improve adenoma detection.

The importance of withdrawal time in assessing quality of colonoscopy became widely disseminated approximately 10-15 years ago. It was apparent that there was a tendency for some physicians to rush the colonoscopy examination, thereby decreasing the likelihood of finding precancerous polyps. The recommended minimum average time for withdrawal of the colonoscope was established at 6 minutes, a threshold which correlated with a significantly increased likelihood of finding precancerous polyps.  My personal average for withdrawal time is 8.9 minutes, which would be expected to result in improved adenoma detection. This is also considered the most important measure of quality of colonoscopy examination at the individual procedure level, ensuring that the examination was not compromised by being unduly rushed.

A significant change in the approach to bowel preparation was instituted several years ago, with patients being encouraged to use a "split-dose" bowel preparation regimen. This means that rather than having patients drink the entire bowel preparation on the day prior to the examination, they are instructed to drink the second half of the bowel cleansing solution on the morning of the procedure itself. By decreasing the time interval between administration of laxatives and the performance of the actual procedure, patients have less time to re-accumulate adherent liquid stool on the bowel wall and have improved visualization at the time of colonoscopy. This greatly increases the likelihood of finding either flat or sessile polyps as well as polyps with serrated pathology. I have utilized a split-dose bowel preparation protocol for over 5 years and have noted dramatically improved visibility at the time of colonoscopy. This has resulted in patients having their preparation rated "good" or "excellent" 94% of the time, improving the likelihood of discovering pre-cancerous polyps.

That being said, no patient has a perfect preparation at the time of their colonoscopy. Adherent liquid stool is especially problematic in preventing detection of smaller or sessile polyps. The tendency is simply to examine the colon "as is" and bring patients back at shorter intervals in order to compensate for less than adequate preparation. It is my opinion that once the patient has made an appropriate effort with bowel preparation and has taken the time and expense to have a colonoscopy performed, it is my duty to "clean up" to the best of my ability, within reason. I typically irrigate the colon with a very generous amount of water in order to remove any remaining adherent stool and thereby further improve adenoma detection.

Holy Cross was fortunate to purchase the newest generation colonoscopes from Olympus approximately 5 years ago, namely the 190 series. There is a specific setting on these scopes called Narrow Band Imaging (NBI) which facilitates the discovery of sessile or serrated polyps. Unfortunately, most physicians utilize this setting very little, if at all. I utilize NBI for the entirety of my colonoscope withdrawal, in order to improve adenoma detection rate and can personally attest that it has significantly impacted my ability to find flat or serrated polyps.

Additionally, there has been a tendency to overuse colonoscopy and have patients return at intervals which are shorter than recommended by professional societies. Many physicians developed a tendency to have patient's return for their "five-year" repeat examination, even in situations where bowel preparation was adequate and there were no precancerous polyps discovered. The recommended interval in such circumstances is 10 years. Additionally, when patients are found to have precancerous polyps, the recommended intervals are frequently overly-shortened. Use of appropriate intervals for follow-up colonoscopies after index examinations lowers the patient's individual procedural risk and also lowers the cost to a health system, overall. I am mindful to provide patients with a 10-year repeat interval for normal examinations if I can confidently do so, unless personal risk factors dictate otherwise. Likewise, I strive to adhere to the recommended intervals for follow-up of precancerous polyps, rather than bringing patients back more frequently than they truly require.

Achieving excellence when taking care of patients does not occur in a vacuum or with a single person, however. I work alongside a very dedicated staff of endoscopy nurses, technicians and unit managers. I am also blessed with tremendously talented and hard-working office staff, many of whom have worked with me for over 10 years and share my passion for providing excellent care. It is my sincere hope that I can continue to do so and that our endoscopy unit will continue to strive to be a leader in outstanding patient care with the support of a hospital that I am proud to be affiliated with - Holy Cross.

Dr. Patrick Amar practices with the Holy Cross Medical Group in Fort Lauderdale. He may be reached at 954-928-1778.


Add new comment

Plain text

  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
  • Allowed HTML tags: <em> <strong> <cite> <blockquote> <code> <ul> <ol> <li> <dl> <dt> <dd>
By submitting this form, you accept the Mollom privacy policy.

About Holy Cross Hospital

Holy Cross Hospital is a nonprofit, Catholic hospital in Fort Lauderdale, Florida, dedicated to innovative, high quality and compassionate care. For nearly six decades, Holy Cross has continuously expanded its services to provide leading-edge care for their patients in Florida and for those from elsewhere in the United States. Holy Cross also offers an International Services program to ensure that patients from outside the U.S. receive the care they need.

Blog Categories

Blog Archive