We're going to close out this session in another area that is near and dear to my heart, which is quality metrics. And if you can't measure it, as we know, we can improve. So Dr. Samir Gupta will speak to us about this. He's a staff physician at the VA San Diego and a professor of medicine at UC San Diego. In addition, Dr. Gupta is the MPI of an NCI cancer moonshot study that's partnering with local federal qualified health centers to study the impact of a regional mailed fit outreach program and structured abnormal test follow-up on screening. Sounds perfect. Thank you. He leads CRC screening and follow-up initiatives at his VA, and he served on the U.S. Multi-Society Task Force on colorectal cancer. Dr. Gupta has his medical degree from the University of Michigan Medical School. Thank you and welcome. Thank you. Thank you very much. All right. So I'm just thrilled the ASGE has supported this. I think this is a really important problem, and we really have an opportunity right now to not only address the abnormal fit, but to get ahead of what I think is a tidal wave of increase in non-invasive tests that's really going to have a big impact on our ASG members. So I think this is really important. These are my disclosures, and I want to talk just a little bit about the rationale for why we might need a quality metric. I will give you my take on proposed quality metrics. I'm really interested to hear about what has already been developed, and I hope there's a lot of overlap, and then just some strategies for promoting and supporting implementation of a quality metric, the things that go into supporting abnormal fit follow-ups, some of which has already been discussed extensively today, so I won't dwell on that. So the rationale. So the public health rationale is that people with abnormal fit are at increased risk. We've talked about this a lot. Health equity non-invasive tests are often more commonly done in populations at risk for non-screening. So if we don't get these people to colonoscopy, we're actually making disparities worse, even though we think we're doing right by getting screening rates up. And then there's a big, I think, wasted resources problem at the clinic system, insurance, and public health level. So for example, if we're going to have more and more non-invasive tests that cost $500 or more and insure society, as a taxpayer, I'm paying for Medi-Cal. If we're going to spend money on this, but we're not getting the colonoscopy, that's completely wasted money. From the primary care practice, I think the moral argument is missed opportunities for prevention, but I don't think we talk enough about liability. One of the biggest reasons for getting sued is not following up on an abnormal cancer screening test. And we need to bring this up more, because it's a real issue. And then from the GI practice perspective, again, the moral opportunity is missed opportunities for prevention. And then I want to underscore this need that I think we need in the next few years to adapt to the changing screening and practice landscape. And here, I would encourage everyone to be prepared to think about two scenarios, an expanded number of stool-based tests with sensitivity that is better than fit, and a blood-based test with performance for colon cancer detection that is similar or better than fit. So these are top-line results that I've found. These are unpublished from 2023. The exact sciences data were already brought up for the Blue Sea study, sensitivity of colon cancer, 94 percent, for advanced lesions, 43 percent, at a specificity of 93 percent. There's the CRC PREVENT study that is a multi-target stool RNA fit test, sensitivity, 94 percent, advanced lesions sensitivity, 45 percent, at a specificity of 88 percent. And then this ECLIPSE study from Garden Health, sensitivity for colon cancer of 83 percent, for advanced lesions, only 13 percent, and then a specificity of 90 percent. So we can sort of take out our crystal ball and just think about, what is this going to mean for practice? And what I think it means is we have an opportunity to get 30 to 40 percent of the population who's not up to date, up to date. There'll be more patients with abnormal tests requiring timely colonoscopy. So the positivity rate for a fit is about 6 percent. But for these tests, they're all going to be 10 percent or more, 10 to 15 percent. So lots more abnormals. The patients and primary care clinicians, I think, may increasingly choose non-invasive options. I don't have a lot of faith that the average primary care doc has enough time to explain the nuances of detection and prevention of cancer through finding polyps if you go for a colonoscopy, versus when you check your cholesterol today, I can just do this blood test to get you up to date with colon cancer screening. And then I think as a result of this, there can be displacement of patients from screening colonoscopy to non-invasive tests. So I really think we as an organization, ASG, should take this as an opportunity to take control of the full screening process so that this really lessens the practical impact on GI practice if this does come to fruition. So the other rationale here, it's clear that colon cancer quality metrics have been the single most important, in my opinion, driving force for improving screening. We saw that with the HEDIS metric and driving private insurance plans to compete on screening. There's the universal data set metric for federally qualified health centers, which has been driving FQHCs to pay attention to their screening rates and try to do something about them. And then what I'm hoping we're going to see with the addition of colon cancer screening to the Centers for Medicare and Medicaid Services adult core set, this is expected to drive improvement in Medicaid-managed care performance because when they start publicly reporting their data, they're not going to like what they see. We're going to see 40%, 50% up-to-date rates. It's not going to look good. So what could be some proposed metrics? And this is very simple that I'll present. The first metric here we could think about is the proportion of patients with an abnormal non-invasive test who complete colonoscopy. So the denominator here is the number with an abnormal test, and the top line is number with colonoscopy completion. And the good thing about this, at least if we think about FIT data, is that increased completion should lead to decreased colon cancer incidence and mortality. There are very few quality metrics that are tied to a decrease in incidence of cancer or certainly for mortality for anything. So I think we should really emphasize why something like this could be important. And the goal of 80% here has been brought up as a placeholder. So we can also think about a second metric. I've heard this brought up in other forums where we could modify the current colon cancer metric to only count complete colon cancer screening. And here we might think about the proportion of patients age-eligible for screening who are up-to-date with screening and abnormal test follow-up. So you can see here the formula gets complex very quickly. So you have the number with an abnormal non-invasive test who complete colonoscopy plus the number with a normal non-invasive test who are up-to-date plus the number who are up-to-date with an invasive test like colonoscopy. And again, the goal would be over 80%. So what could be some trade-offs of this? So I think as we think about the first metric, the pros here, very clear messaging, the laser focus on the biggest challenge, which is abnormal test follow-up. It's tightly tied to mortality, at least for fit. And it's hard to game this. The cons here is it requires a new metric rather than modification of an old metric. How do you set the sampling and follow-up time frame? So do you look, for example, at the first six months of the year? These things can be worked out. And then small numbers in some settings can make measurement of this hard. Like if you just look at an individual primary care clinic, maybe they only have 100 abnormal results in a whole year. When I think about the metric number two, it does promote the idea that screening is a process. So I like that. But the downside or con is that health insurers or systems kind of get away with a minimal focus on the topic. And I want to just emphasize why that could be the case. So if we think about a hypothetical population of 10,000 people, and let's say 40% are up to date with colonoscopy, 30% get up to date with fit, you have 5% abnormal and 50% completion. So if we look at the first metric, if you have 150 with abnormal fit, only 75 will have follow-up colonoscopy. And you have this really bad-looking completion rate of 50% on metric one. But if you use metric two and you do the math, you basically end up with having 69.25% up to date with screening. And then the abnormal test follow-up issue becomes a rounding error. It really will not look important at all. So this is what I'm worried about if we go with modifying the current HEDIS. So just a few thoughts about implementation here. In terms of feasibility, some people might bring up, this is not feasible. It actually is feasible. We have multiple use cases. We've heard cases today. The Kaiser example presented earlier. In San Diego, I can tell you that as part of our access project, the three federally qualified systems that we've had, none of them had trouble with producing lists of patients with abnormal fit that could be worked. So that gives me some faith. We can do that in my university system. We've seen that with work that Ma and Rachel have done at the San Francisco General Hospital. This can be done on a national basis. This has been going on in the VA for so long. Jason has led this. Every VA has access to the list of patients with abnormal fit. We have their name, medical record number, their abnormal test date, and their consult status in terms of colonoscopy or GI consult. And I can tell you at VA San Diego, we've been working this list. And we've reduced our wait list from 300 to 100 over a six-month period by having a GI navigator work the list. We trained the navigator in how to use the registry and using a toolkit, just a slide set that we gave them, on promoting follow-up and also allowing for some Saturday endoscopy sessions where we prioritized scheduling of abnormal fit patients first. And I can tell you, even out of the 100 that are still waiting, 30 are abnormal fits in the last month or so, 30 are probably patients who are never going to do it. We're having trouble finding them. So I think we've been able to do really well with this. In terms of enabling change, so we've talked a little bit about what does it take. So you have to have the care coordinator or navigator. And I think the extent to which we can leverage things going on right now with Medicare policy around navigation or the extent to which we can convince GI practices that it's in their interest, their financial interest to invest in helping get the primary care patients through their doors could be really helpful. You need some kind of tracking mechanism, a registry, or even a spreadsheet. Ideally, you're evaluating process measures such as referral, scheduling, and completion by clinic. In terms of toolkits, you need some kind of checklist with key steps, strategies for addressing common issues like inappropriate repeat fit. We've heard about that. Explaining away abnormal results, also a problem. And then, of course, colonoscopy barriers. Some kind of standardized queries for these data and standardized workflows. So at San Francisco General Hospital, Ma and others have reported on making sure that they know who is responsible, adding the abnormal fit to the problem list, talking about standardized telephone encounters in which they interact with the patients using dot phrases, and leveraging structured data elements and reporting to get attention to this. And then, again, just to circle back, I think we need strategies for primary clinician engagement that this is both a patient health issue and a major liability issue for them. Again, just thinking about more non-invasive tests, higher percentage positivity, they're going to have a bigger burden of this. So in summary, I think as far as the rationale for an abnormal screening metric, I think we need this to improve public health and optimize the promise of non-invasive tests, to reduce liability, to optimize the role of the GI in colon cancer screening, and reduce the disruptive impact of non-invasive screening on GI practice. And to me, this would be the most powerful action to enable change. I would argue that the best metric is proportion of people with an abnormal test who complete colonoscopy. I think that measurement is feasible, and it's poised to make a difference, so thank you. Oh, and I just want to mention, a lot of my thinking on this has really been guided by Ma, who's a thought leader, and I'm really grateful to him for sending me some of his slides and also looking at my presentation, so thank you.

Dr. Samir Gupta

quality metrics

healthcare

non-invasive tests

colon cancer detection