And now it's my absolute pleasure to introduce Dr. Jason Dominitz, Dr. Dominitz, who will address quality indicators for colonoscopy. Dr. Dominitz doesn't really need introduction, but he's the national director of gastroenterology for the VA Health Administration and the acting chief of gastroenterology for the VA Puget Sound, as well as a professor of gastroenterology and medicine at the University of Washington. He, among multiple other roles, has served as a member of the ASGE Quality Assurance and Endoscopy Committee and currently serves as a member of the GI Quick Colonoscopy Measures Subcommittee and has really taught all of us a lot about performing high-quality endoscopy. So Jason, I'll turn it over to you. Thank you. Thank you so much, Dr. Pannulla, and thank you, Dr. Yang and ASGE, for inviting me to speak with you all this morning about quality indicators for colonoscopy. I hope you're all having a good morning today. It is a little early for me on the West Coast here. I have no disclosures. Now for the sake of time, I'm going to not go through the learning objectives and the outline. Of course, you guys will have access to GILeap and have access to all of these slides. Of course, all of our patients deserve high-quality colonoscopy, and we want to make sure that we are inclusive and equitable about that, as Dr. Pannulla was just talking about. One point I want to emphasize is that we do have different benchmarks based on gender. You heard about the adenoma detection rate, and we're going to dive into that a little more deeply. But I think it's interesting, if you compare endoscopists, you might see that a female endoscopist has a lower adenoma detection rate than a male endoscopist, and some people may assume that that means that the female endoscopist is not as good. But we know that female endoscopists tend to attract more women patients, and women have lower prevalence of adenomas, and therefore you would expect that would be lower. So despite a lower adenoma detection rate, they actually could be a superior endoscopist. Now quality and colonoscopy, you heard from Dr. Pannulla about why it matters. Colonoscopy is the most common endoscopic procedure in the U.S. You heard about the data on colon cancer being affected by the quality of the endoscopist whose variable performance and that variation impacts the effectiveness. We know that screening colonoscopy can reduce colon cancer incidence and mortality. We have now randomized controlled trial evidence to support that just in the last year. But we want to make sure that everybody is getting access to high-quality colonoscopy, as I said a few minutes ago. And so this has become a priority for the ASGE to help make sure everyone's embracing that. And kudos to all of you for attending this course today. You heard a bit about this in one of Dr. Pannulla's slides. The proportion of cancers that are found in people who had a colonoscopy that missed it in the last three years is not a huge number, but it is a surprising number. Between about 3% and 9% of all colon cancers that are found are found in someone who had a colonoscopy in the last three years that apparently missed it. Why does this happen? Well, it could be that the endoscopist missed the lesion, that there was a polyp or a cancer that was not seen but was present at that time period. And in this study from Doug Robertson and colleagues, they looked at thousands of patients undergoing colonoscopy and followed for four years. And they estimate that 52% of these cancers were probably missed lesions. Another about 20% were probably lesions that were incompletely removed at the time of colonoscopy. And then only about 24% are thought to be new lesions that grew rapidly in that interval since the last colonoscopy. So that should give us some pause as we think about how well are we performing colonoscopy because we certainly don't want our patients to leave our endoscopy unit, you know, getting a, you know, being told that the colonoscopy looks good or we removed all the lesions. And then within a few years, they're being diagnosed with colon cancer. That is just something we cannot really accept happening. You know, we need to make sure we're doing a better job and not missing those lesions or incompletely removing them. Let's start with some poll questions. First one, which of the following is not a priority quality indicator for colonoscopy? So you've got five choices. Which one is not one of our priority indicators? The frequency of documenting an appropriate indication, the frequency of appropriate surveillance recommendations, the frequency with which the bowel prep is deemed adequate, the frequency of sequel intubation with photo documentation, and finally, the frequency of adenoma detection or adenoma detection rate. Okay, so yeah, this is a hard one. So the appropriate indication actually is the one that got the best response, but that actually is a priority indicator. So let's go through these. And so it's good that we're talking about this. I know it's a tricky question. Let's see, I don't have the ability to move forward. There we go. So there are several documents that were published back in 2015, and Dr. Panalla referred to them on one of his slides. These documents are actually undergoing an update right now. I'm part of the team working on the colonoscopy quality indicators. But there are quality indicators for all endoscopic procedures, and then there's specific quality indicators for procedures like colonoscopy, ERCP, EGD, et cetera. And when you look at the quality indicators common to all procedures, we have three priority indicators as shown on the slide. The first one is the appropriate indication quality indicator. We do want to make sure that 90% of the time when we're scoping, we're scoping somebody for a good indication, and there is a document that the ASG Standards of Practice Committee puts out that lists the appropriate indications. The other two priority indicators are appropriate use of prophylactic antibiotics and management of endothrombotic therapy being documented before the procedure. And I think we can all agree that these are important things for all endoscopic procedures. Now you'll note at the top of the slide that there are a lot of quality indicators. There's nine pre-procedure, five intra-procedure, nine post-procedure. I'm not going to go through those for the sake of time, and I think I would bore you to tears if I went through them line by line. When we look at quality indicators for colonoscopy, we have these three priority indicators. The adenoma detection rate is first. The second is following appropriate surveillance recommendations or the 10-year interval for screening colonoscopies in average-risk people with a negative exam. And then third is the sequelae intubation rate. So again, the priority quality indicators are appropriate indication, the adenoma detection rate, the appropriate surveillance recommendations, and sequelae intubation rate. Now, as I said earlier, there's a lot of quality indicators for the general quality indicators for all procedures. For colonoscopy, we have a lot more. So there's over 30 quality indicators for colonoscopy that you could use. But I think it's important to focus on the priority quality indicators. And once you've got a good handle on those, then you can think about the other ones as well. Now, here's just an example of what's in the document. Again, I'm not going to go through these line by line, but the document gives you the grade of the recommendation, how good is the evidence, whether it's a process measure or an outcome measure. And you heard from Dr. Pinella about the Don and Budin model. And we want to focus on outcome measures as much as possible. But those are relatively uncommon because outcome measures are harder to measure. And so we often focus on process measures. And then we have the performance targets. So now let's talk about some of the surveillance recommendations. Let's do a poll. When should you recommend surveillance in an average-risk individual found to have an 8-millimeter adenoma and a 9-millimeter adenoma on a high-quality colonoscopy? In other words, this person does not have a family history of cancer, they don't have a prior history of adenomas, you find these two polyps, 8 and 9 millimeters, on high-quality colonoscopy. No penalty for wrong answers. I'm stealing Eden's line, sorry, Eden. Okay. So you've got two adenomas that are both under a centimeter. And so the guidelines are actually 7 to 10 years. And so let's talk about this a little bit. And I don't expect everybody to know all these answers. Let's do another one. We have a patient, the same patient, let's say, instead of those two small polyps, they're found to have a 10-millimeter adenoma on the high-quality colonoscopy. One year, three, three to five, five, five to 10, what do you think? And there's what people are thinking, Dr. Coleman. And three years. And that, very good. That is the correct answer. This is a high-risk lesion. This is an advanced adenoma. Any adenoma a centimeter or larger, or if there's villicystology or hygrodysplasia, it's considered an advanced adenoma and three-year follow-up is appropriate. So let's see. Let's move on. So there's a lot of underuse and overuse of surveillance. Now this data I'm showing you is from Rocky Schoen's PLCO study. This was a study of screening sigmoidoscopy that was done some years ago. And they followed patients who had adenomas or no adenomas and high-grade or advanced adenomas and looked at what happened in terms of colonoscopy over time. And what I'm highlighting here is that among people with no adenomas, this is the black line on the bottom, with no adenomas, at five years, 26% had had another colonoscopy already. Now when there's no adenomas, we would recommend 10 years for the next colonoscopy. Now certainly some of these people may have developed symptoms, something may have happened that warranted another colonoscopy, but 26% seems like a high number. So that is evidence of overuse of surveillance. And we know that a lot of doctors do recommend a five-year follow-up exam after normal colonoscopy. Now when you have advanced adenoma, we just said you should have a three-year follow-up. And when we looked at patients in PLCO who had advanced adenomas, they should have come back by three years. We would hope that that dark black line would be close to 100% by three years, but it's only 58% at five years. So two years later, this is underuse. These patients are at risk for developing an interval cancer, and if they don't come back, we lose that opportunity to remove a lesion before it becomes cancerous, or to detect it when it's still an early, more treatable stage of cancer. So the U.S. Multi-Society Task Force, which is made up of members of the ACG, ASG, and AGA, put out the Multi-Society Task Force guidelines on a number of aspects of colon cancer screening and surveillance. The polyp surveillance recommendation is updated in 2020, and there is this figure that I think everyone should have posted wherever they're making decisions about follow-up. My endoscopy unit has the poster in every procedure room, and in the fellows room where the fellows are often making decisions, and doctors have it in their office, maybe as a, you know, 8 1⁄2 by 11 piece of paper. There is a poster in the gastro version, that's what's shown here on the reference, it's called Spotlight, that you can actually print out that poster, it came in the issue of gastro, and it has this figure that shows you what to do, and you can see on the left hand side, with a normal colonoscopy, or less than or equal to 20 hyperplastic polyps under a centimeter, you do 10-year follow-up. That first poll I had on this topic, one or two adenomas under a centimeter in size gets you 7 to 10-year follow-up. If you have one or two small cystic ulcerated lesions, it's 5 to 10, etc. There are some major changes from the earlier version of this guideline. We used to recommend a 5 to 10-year follow-up for those patients with the two small adenomas, and it's been stretched out to 7 to 10 years, or I should say delayed to 7 to 10 years, because these patients have a similar risk of cancer to those with a normal colonoscopy, and they have a lower risk of colon cancer compared to the general population. So we probably were using more colonoscopy than we needed to. If you go back to when I trained as a fellow, people were bringing patients back every one to three years. Anytime they found an adenoma, one of my attendings said it was like an annuity, because he could help make his boat payments with every adenoma he found. But now we know that a lot of these lesions are low-risk, and especially as we are using higher-definition equipment, we're getting better at cleaning the colon, we're getting better at examining the colon, and we're finding more and more tiny lesions. Now if you find two, you know, 4-millimeter, 5-millimeter, 6-millimeter adenomas, I'm perfectly comfortable waiting 10 years. If I find two 9-millimeter adenomas, and I know some people believe there's no such thing as a 9-millimeter adenoma, but if I find two 9-millimeter adenomas, I might prefer to go with a 7-year follow-up. We also reinforced in those guidelines the importance of surveillance colonoscopy at three years for those with advanced adenomas, as I was saying before on that PLCO study. We really want these people to come back, because they are at increased risk for colon cancer. These are the people at higher risk for interval cancers. We want them to come back. Their risk of colon cancer is higher than the general population. We want to look for a polyp that may have been missed, or a polyp that may have been incompletely resected, or a new lesion. We used to recommend that anybody with three or more adenomas comes back in three years, but as I said a few minutes ago, we're finding more and more small lesions as we're using better scopes and better technique. The risk of patients with three or four adenomas is not as high as we used to believe. Because of that, you can now stretch the interval out to five years instead of bringing everybody back at three. If you found four 9-millimeter adenomas, I might be bringing the patient back in three years. But if you find four 3, 4, 5-millimeter diminutive lesions, I don't think you need to bring them back so soon, and you can wait five years. The same poster I mentioned before also has this figure, which is one of my favorite figures, because I can never remember this. I helped write this guideline, and then I still look at this figure all the time, or I should say table. What this does is, on the left-hand side, you have the baseline findings, so that patient who had the 8-millimeter and 9-millimeter adenomas is here in this cell. And then we recommended that 7- to 10-year follow-up, and now they've come back for their first exam. Now, if it's a normal exam, they can come back in 10 years. People used to bring everybody back in five years if they ever had an adenoma. You don't need to do that. You can wait 10 years. One or two adenomas, again, the 7- to 10-year, the interval if they're small. So basically, these recommendations are the same as the first exam. Now, you can see, if you go to the high-risk lesions here, the bottom row, we have that three-year follow-up. If it's normal, now they're on the five-year interval. We don't know exactly what we should do the next time, so you do that second surveillance. If it's normal on the first surveillance and normal on the second surveillance, do you need to bring them back in five years? There's no table that tells us what to do there. You have to use your clinical judgment. So, there are a lot of intra-procedure quality indicators. For the sake of time, I'm not going to go through them all, but I bolded the priority indicators, which we've talked about before, the sequent debation rate and the adenoma detection rate, and we'll go into a little more detail on those. So, if the colonoscope doesn't reach the CECM, assuming the patient has the CECM, there's no prior surgery, then you're not going to be able to clear the patient of polyps or cancers. So, you really need to get to the CECM. Low sequent intubation rates are associated with higher interval proximal cancer detection rates, and I'll show you that study in a moment. And photo documentation is mandatory. You want pictures of the appendiceal orifice and the IC valve, as you saw in the prior slide. If you're not sure you're seeing that, and sometimes it's not as clear as in other patients, go in the TI and take a picture like you see in the upper right-hand corner. The target is to hit the, you know, to get photography of the CECM in 90% of all exams and 95% of screening exams. You can exclude from the quality indicator those patients with a poor PrEP or severe colitis, or when you weren't intending to go to the CECM, for example, if you're doing a follow-up on a, you know, a piecemeal resection of a transverse colon polyp. Here's a nice study from Canada from Nancy Baxter and colleagues, where they, in Canada, they code the depth of insertion. So they're able to identify the patients who did not have a complete colonoscopy. And when providers had below an 80% sequent intubation rate, the risk of their patients having post-colonoscopy colorectal cancer was higher. Here we see that for endoscopists with a sequent intubation rate at 85% to 89%, the patient's risk of post-colonoscopy CRC in the proximal colon was reduced by 31%, 34%, and then 28% as the sequent intubation rate improved. So we know that it's important to get to the CECM. It makes sense. How about the adenoma detection? Dr. Pinella spoke about this briefly. It's the frequency with which adenomas are detected in asymptomatic average individuals over age 50 who are undergoing a screening colonoscopy. So you're supposed to only look at diagnostic and surveillance colonoscopies, sorry, you're only supposed to look at screening colonoscopies and exclude the diagnostic and surveillance exams. This is our single most important quality metric in colonoscopy. It has a direct correlation with colon cancer, and we have a target of 25% overall, 20% in women, 30% in men. I work in the VA, where about 90, 95% of our patients are men. And therefore we expect our doctors to get a sequent adenoma detection rate of at least 30%. In fact, our average adenoma detection rate in the VA is well above that, as it is in GI-Quick. You heard mention of GI-Quick, which is a data registry that the ASGE has created in partnership with the ACG. And if you're not a member of GI-Quick, I suggest you consider it because it's a nice way to track your quality metrics and benchmark them against other practices. There's over 700 practices currently using GI-Quick. In GI-Quick, the adenoma detection rate average for screening colonoscopies is 39%. So this benchmark of 25% is really low. And you saw data from Mayo Clinic, you saw that in general the average was well above 25% in the Mayo Clinic. You saw this slide earlier from Dr. Pinala. This is Kaiser data, Doug Corley and colleagues. At the top, I have the adenoma detection rate clintiles for their physicians, over 100 physicians. And the highest adenoma detection rate physician is 53%. The lowest is 7.5%. So a lot of variation between physicians. And what you see is that the risk of cancer after colonoscopy decreases as the physician's adenoma detection rate increases. Again, as Dr. Pinala said, every 1% increase in ADR is associated with a 3% decrease in the risk of cancer and a 5% decrease in the risk of fatal cancer. So this is our best quality indicator today. Now, withdrawal time is also a metric that's been touted to track colonoscopy quality. And some people have tried improving colonoscopy quality by demanding that physicians meet a set withdrawal time. The studies doing that have been mixed because of what has been sometimes referred to as the rectal waiting time. A physician does their colonoscopy, they get to the CECM, they start pulling the scope back, they get to the rectum and they say, nurse, how long have I been withdrawing? And the nurse says, you've been withdrawing for five minutes. So the doctor says, okay, I'll sit here for a minute to get to the minimum required six minute time. And that doesn't work. That doesn't improve the quality of the exam. You need to spend the time examining the colon for polyps. And what you see here from Dr. Shawkat of NYU, the study she did when she was in Minnesota, 51 gastroenterologists from a practice in Minnesota looking at their average withdrawal time when no interventions are done. So no polypectomies or biopsies. And you can see that here we have a physician whose average annual withdrawal time is around 15 minutes. And as the time is over nine minutes, we see that the interval cancer rate on the y-axis is fairly low. As the time gets shorter, the interval with cancer rate takes off exponentially where the physician who has a four minute average withdrawal has three times the interval cancer rate of the physicians who have spent nine minutes or more on average. So we know that this is a surrogate for the quality of the exam. Retrospectively, it works great. Prospectively, when you try to get people to change the withdrawal time, it doesn't impact quality unless you're spending that time fruitfully examining the colon. Now we have a number of post-procedure quality indicators largely related to complications. Fortunately, complications are fairly rare, but making it hard to compare physicians and see if they are outliers. But of course, it still is important to track complications and it can give you a signal if there's an issue with a physician or their practice. So how do we improve colonoscopy quality? Well, Dr. Pinella mentioned this, just simply giving people a report card improves their quality. Here's one of the landmark studies from Dr. Kahi of Indianapolis, where they had high adenomatotection rates. Look at this. Before the intervention, they had a 45% adenomatotection rate, but they gave people a report card and they got it up to 54% adenomatotection rate, and proximal adenomatotection increased significantly. And the sequel intubation rate started at 96%, which is above the benchmark of 90%, and they got it to 98%. So remarkable improvements in performance, really incredible quality. From the Mayo Clinic, Mike Wallace and colleagues did the equip study where they did training of how to do colonoscopy. It's not simply, you know, pulling the scope back. You need to deflect the tip, flatten the folds, wash, distend. And with that training and training and how to recognize subtle lesions, they took people who had, you know, high adenomatotection rates and made them even higher. We have a number of tools that are disposable at our disposal to help us with improving adenomatotection rate. And there's also techniques. There's a paper from Asma Shalkut and colleagues. I'm not sure if the reference is showing up on your screen. It's not showing up online, but Asma Shalkut is the first author. It's from the ASG Quality Committee on techniques to improve adenomatotection techniques and tools. And I urge you to look at that, especially if there are any issues with colonoscopy quality in terms of adenomatotection rate in your practice. Now, a new wave of tools are arriving. You may have heard of computer-aided detection or CAD-E, an artificial intelligence. What you see on the screen here is a green bounding box that a computer device puts on the screen to highlight a polyp that is somewhat subtle. Physicians might miss this. So CAD-E and AI are available. There are, I believe, three FDA clear devices. There's one that's been around for a couple of years. There's another one that was approved last year, and one was just approved by the FDA in the last week or so. And some practices are using these. So that is a little bit mixed. Initial studies show up to a 14% improvement adenomatotection rate. There's some studies like from Yuri Ladovim at Stanford showing that there was no improvement, and one from Israel. And then there's some others showing some benefits. So I think we're looking at this data in the VA. We outfitted 43 different VA facilities with AI, and we're looking at that data now, and we're writing up a manuscript. Stay tuned for this. If you're not, if you have trouble with adenomatotection rate at your facility, I think this is the future. There will be more use of AI. Raj Goswami from Northwestern has a nice paper in gastroenterology on best practices in colonoscopy and how to improve the quality. We're going to talk about bowel prep a little later this morning, about using split prep, using high depth scopes, performing a second look in the right colon, using cold snare to get complete reception of polyps, and following surveillance recommendations. All of these issues are discussed. I think it's a nice paper. If you want to make sure you have a high quality endoscopy unit, I suggest you read that paper by Shalkat et al and this paper by Goswami. And as Dr. Penala spoke about, you have the Plan-Do-Study-Act or the Six Sigma approach to quality improvement. It's a multi-component process. You've got to start looking at your data and then come up with interventions. Could be techniques, could be technology to try to improve the quality of colonoscopy at your facility. There are some challenges in quality improvement. However, Dr. Penala also touched on this. It's hard to measure these issues. In the VA, we surveyed our endoscopists around the country, and most facilities were tracking bowel prep quality, cecal intubation rate, and complications, but only about 60% were tracking surveillance intervals and withdrawal time and adenoma detection rate. And so we've been working in the VA to improve that. We've strengthened some of our requirements and we've built tools to automate some of these processes. The adenoma detection rate, as I said, is our best overall quality metric. It can indirectly reflect other issues such as prep, withdrawal time, and technique. Now, some people will exclude poor prep and incomplete colonoscopy from the adenoma detection rate. That's not clearly defined in the quality document, and we will be updating that in the new document that is coming out in the not too distant future. But it does take a large number of screening colonoscopies per provider to get narrow confidence intervals on the adenoma detection rate. The paper that's in press from Gastroenterology, Doug Corley's group, has looked at using the overall adenoma detection rate on the far left rather than the screening adenoma detection rate. You can see that they're fairly similar. There is about a 6% increase in the average value when you use overall ADR, because surveillance colonoscopies find more polyps than a screening colonoscopy. But when they correlate the quartiles, like the quintiles I showed you before, based on overall ADR and the black line, or screening ADR on the red line with the post-colonoscopy colorectal cancer risk, they overlap almost perfectly. So this is a really good quality metric. This is what we use in the VA, the overall adenoma detection rate. It's a lot easier to measure, and you get narrower confidence intervals. And the new document coming out from the ASGE and ASGE is considering this kind of change. I can't speak to what the end result will be, but I would not be surprised if in the future we have different, that the ADR is defined differently, or maybe you have multiple different definitions of ADR that can be used. So it might not just be screening, and the benchmarks may rise. I suspect the benchmarks will increase above that 25%. Now, the ADR does not include CESL serrated lesions. There may be a separate measure for that. That's something under consideration. It can be labor-intensive to track adenoma detection rates, to do it manually. If you use GI-QUIC, it does report it for you, but you've got to enter the data. You've got to enter your path results. Some places have automated that, and we've done that in the VA. It does emphasize the detection of small lesions, and there is this risk of one and done. For example, if you get to the CECM and you find a sequel adenoma, you take it out, are you going to try as hard to look for adenomas as you pull back, or you might just find that one and then pull the scope back quickly? Now, I don't think anybody would consciously do that, but if the adenoma detection rate is being tracked, people might relax once they find an adenoma. If you track something like the adenomas per colonoscopy, which is the count of all adenomas in a colonoscopy, that would alleviate that concern. However, tracking adenomas per colonoscopy is extremely challenging and problematic. To do it accurately, you'd need to put every polyp in a separate chart. That would get to be very expensive for the patients in the healthcare system. Now, surveillance recommendations, tracking that is likewise challenging. You need to manually review the chart to make sure you've got the right indication, the bowel prep quality, the depth of insertion, the findings, and then what the recommendation was, which requires linkage to pathology. So, you can't do that at the time of the procedure. Uri Ladebaum has come up with a semi-automated process, basically looking at the pathology letter, asking the physician when they tell the patient what the bowel plan is to say why they're saying that plan, and you can use an algorithm to see if they're making the right recommendation based on the guidelines. We have developed that same process in the VA so that you can see which physicians are following the guidelines and which aren't. In my facility, we've been doing manual chart review. We pick five charts every six months for every physician and review them to see if they're making appropriate surveillance recommendations. It's eye-opening to see what your colleagues do. Some of them are way off from the guidelines, and then you can talk to them about it and try to rectify that process. But if you don't measure it, you can't do anything about it. And then, physical intubation rate, you need to look at the – you can't just go by self-report alone. You need to look at the photographs. And again, I told you we review at least five charts, usually 10 charts every six months. And I look at the photographs and see, is there a CECM documented? I believe in the future, we will have the computer doing this for us. You know, we've got great facial recognition software. Your phone recognizes your face. Why can't your computer tell you this is the CECM? And there is actually some data. People have developed that, but we need to get that integrated into our endoscopic software like Probation, EndoSoft, EndoPro, et cetera. Despite all these challenges, measurement of colonoscopy quality is really the standard of care today. Some places even publicly report adenoma detection rates. You need to develop workflows to facilitate quality measurement, and using endoscopic software with reporting tools is really quite helpful. Using GI Quick, I think, is a great process for you to consider. Audit the photo documentation of CECM intubation and audit surveillance recommendations. So to conclude, colonoscopy quality is associated with cancer incidence and mortality. Reporting the quality will lead to improved quality. And you need to focus on the four quality indicators that are priority indicators, as I mentioned. Appropriate indication, adenoma detection rate, CECM intubation with photo documentation, and appropriate surveillance recommendation. Low-level adenoma detection creates a significant risk for our patients, and we have many proven interventions, tools, and techniques that can help us improve the adenoma detection rate. If you don't measure it, you don't know if you've achieved it. Thank you very much for your time and attention.

quality indicators

colonoscopy

adenoma detection rate

cecal intubation

surveillance recommendations

artificial intelligence tools

patient outcomes