Colonoscopies are an obvious target in the current push to decrease health care costs, given their expense and utilization, said experts at Digestive Disease Week (DDW) in Chicago in May.
“Colonoscopy has been really good to our specialty. It is the goose that has laid the golden egg,” said David A. Lieberman, MD.
But innovations in payment systems and technology are likely to change this situation soon. “There are a number of potentially disruptive technologies that may affect utilization of colonoscopy in the future,” said Dr. Lieberman, a professor of medicine at Oregon Health & Science University in Portland.
Potential paths to a higher-value colon cancer screening system include developing alternative screening technologies, refining risk assessment and surveillance, and lowering the cost of sedation.
Other ways to screen
Several researchers presented new data on the most widely used alternative to colonoscopy, the fecal immunochemical test (FIT). More than 200,000 people were offered FIT in a study by Kaiser Permanente Northern California (KPNC), and the results were compared with colonoscopy. The study found an average sensitivity of 82.1% for FIT over the 4 years of annual testing.
“Although screening with FIT misses some cancers, missed lesions are more likely to be subsequently detected while still in an early and treatable stage,” said lead study author Pauline Mysliwiec, MD, a gastroenterologist and researcher with KPNC.
A similar study in Hong Kong, also presented at the meeting, invited men and women between 50 and 70 years of age to colon cancer screening and allowed them to choose between 4 years of FIT or a single colonoscopy. About 5,800 participants chose FIT, and 4,800 chose colonoscopy. Again, FIT found about the same rate of cancer as colonoscopy (0.2% vs. 0.3%), but far fewer adenomas.
FIT poses problems as a screening method, according to Dr. Lieberman. “The sensitivity is not high enough that we can do this infrequently,” he said. More frequent testing results in lower adherence, and the lower rate of adenoma detection means that fewer abnormalities may be found at a curable stage. “Perhaps FIT is going to have some limitations when it comes to cancer prevention,” Dr. Lieberman said.
Blood tests, which look for DNA markers of colon cancer, have the same limitation, according to the evidence collected so far, including research presented at DDW. “If they only pick up late-stage lesions, they're not going to be very helpful for us,” Dr. Lieberman said. “Although I call serum markers the Holy Grail of screening ... I don't think we're there yet.”
Computed tomography colonoscopy detects lesions quite well but has other barriers. It's not cheap, it exposes patients to radiation and incidental findings, and there's at least as much bowel prep as with standard colonoscopy. “If the patient cannot go immediately to a colonoscopy after a [CT] finding that would lead to colonoscopy, that means patients have had 2 bowel preps,” said Dr. Lieberman.
Capsule colonoscopy, in which a patient swallows a capsule that captures images of the colon, also requires more thorough prep than a regular colonoscopy, since there's no opportunity for more cleaning during the procedure. “At this point, I don't think it is prime time for screening,” Dr. Lieberman said, noting that capsules may be useful as a follow-up for incomplete colonoscopies or an alternative for patients at high bleeding risk.
But 1 new screening method is likely to hit prime time soon. Developers of a stool-based multi-target DNA test look forward to potential FDA approval later this year, according to Barry Berger, MD, chief medical officer of Exact Sciences Corporation.
Dr. Berger reported on the most recent evidence for the stool test, which was also published in the April 3, 2014, New England Journal of Medicine. The almost 10,000 participants collected a stool sample at home and then followed a simple procedure to stabilize the sample before mailing it back for hemoglobin and DNA testing.
“The multi-target stool DNA approach was significantly better than FIT for every analysis and subanalysis,” said Dr. Berger. “With a single test, 94% of surgically curable colorectal cancer (American Joint Committee on Cancer stage I and II) is identified.” The DNA test identified significantly more advanced adenomas, high-grade dysplasia, and sessile serrated polyps than FIT. The test appears accurate enough that it could potentially be done only every 3 years, as suggested by the American College of Gastroenterology's 2008 guidelines, he added.
The results so far are promising, agreed Dr. Lieberman. “It probably doesn't need to be performed as often [as FIT], but we still need some data to help guide us on frequency,” he cautioned. The cost-effectiveness of this screening method will also be important and will depend on how much the test costs when it hits the market, he noted.
Despite its merits, the stool DNA test “is not a replacement for screening colonoscopy,” even according to Dr. Berger. The test would be an alternative option for patients at low or average risk, especially those who can't or won't get a colonoscopy, he said. Diagnostic colonoscopies would also still be necessary to follow up positive results.
Figuring out what puts a patient at low, average, or high risk of colon cancer was another hot topic at the conference.
“The great majority of people who get screened [for colon cancer] derive no direct benefit from it. These resources could be well used elsewhere,” said Thomas F. Imperiale, MD, FACP, a professor of medicine at Indiana University in Indianapolis. “What we need is a system for risk stratification.”
Dr. Imperiale is working to develop such a system. At Digestive Disease Week 2013, he presented an initial model for predicting risk in 50- to 80-year-olds, which found 5 factors that predicted whether advanced neoplasias would be found in the patients' first colonoscopies: age, sex, a first-degree relative with colorectal cancer, tobacco use, and waist circumference. “The only factor [we investigated] in this short model that was not statistically significant was family history,” he said.
At this year's conference, Dr. Imperiale presented data on a new strategy that applies this short model initially, followed by an expanded prediction model, which added marital status, education, alcohol, NSAID and/or aspirin use, metabolic syndrome, red meat consumption, regular activity during the past 10 years, and vigorous activity within the past year.
In a test of this sequential strategy, 55% of more than 2,500 studied patients were categorized as low risk (and they were found to have a 4.4% risk of advanced neoplasia), 22% were intermediate risk (with a 6.1% risk of advanced neoplasia), and 22% were high risk (with a 24% risk of having advanced neoplasia).
Five distal cancers were found in the low-risk group (779 patients), and 71% of the 65 advanced neoplasms (including the 5 distal cancers) would have been detectable with sigmoidoscopy, leading Dr. Imperiale to conclude that combining the sequential model strategy with sigmoidoscopy might be an effective screening alternative for low-risk patients. Similar results were found in the validation group. “This strategy could improve the efficiency and uptake of colorectal cancer screening,” he said.
Another study presented at the meeting also analyzed risk factors that may be used to target screening colonoscopy, particularly family history. “Family history of colon cancer is the only risk factor that leads to screening at an earlier age,” said Philip Schoenfeld, MD, associate professor of medicine at the University of Michigan in Ann Arbor.
He and his colleagues compared adenoma prevalence in patients who had a first colonoscopy between ages 40 and 49, either due to a first-degree relative with colon cancer or due to an indication unrelated to colon cancer, such as diarrhea.
“Individuals with a family history of colon cancer did not have an increased likelihood of adenomas of any size compared to the controls: 28.4% versus 25.4%. For advanced adenomas, there was no difference in the prevalence: 4.6% in individuals with a family history versus 5.1% among control patients,” reported Dr. Schoenfeld.
The patients also filled out questionnaires about other potential risk factors, and according to their responses, male gender and current tobacco use increased risk of adenoma, while regular aspirin use decreased it.
There are 2 ways that such findings, if confirmed, could be used to refine screening recommendations, according to Dr. Schoenfeld. Patients with family history of colon cancer who are low risk according to other factors may not need to start screening at age 40. “Alternatively, you might say that if you have morbid obesity, ongoing tobacco use, male gender, maybe we should start screening earlier,” he said.
Another debate in colonoscopy utilization is once screening is started, how often should it be repeated? About a quarter of colonoscopies done today in patients over age 50 are for surveillance, the experts said. “We have to acknowledge that there is some overuse of colonoscopy,” said Dr. Schoenfeld. He reviewed data showing that significant percentages of patients receive repeat examinations in shorter intervals than recommended by guidelines.
One cause is inadequate bowel preparation, and his solution was simple. “Splitting the prep is the single intervention that's going to have the biggest impact,” Dr. Schoenfeld said, noting that 4 to 6 hours before an exam has been shown to be the optimal time to take the second dose.
The optimal time for colon cancer surveillance has been set by guidelines but not so definitively proven by the evidence, according to Dr. Lieberman.
“Does surveillance reduce colon cancer mortality, and does surveillance reduce colon cancer incidence?” he said. “The answer, if we're honest, is that we do not know. That may be a shocking answer. Perhaps that primary benefit is derived from the identification of polyps and removal of polyps on the initial, high-quality baseline colonoscopy.”
He reviewed evidence showing the long protective effect of an initial colonoscopy, as well as data showing that patients with a couple of very small adenomas on an initial exam—who are guideline-directed to return in 5 to 10 years—may be as low risk as those with no findings.
“New information on surveillance programs now suggests that many patients can have longer intervals,” Dr. Lieberman said. Thus, he said, the focus in colonoscopies should be on higher quality, lower quantity. “Payers are going to expect us to provide high-quality exams, and they're going to expect us to show it. They're going to expect us to provide exams at intervals in accordance with guidelines,” he said.
Payers are also taking a close look at a major component of colonoscopy cost: sedation. “Routine use of intravenous sedation is basically the rule in the U.S., with nearly 99% of patients undergoing intravenous sedation,” said Jason Dominitz, MD, a professor of medicine at the University of Washington in Seattle.
Yet research, and standard practice in other countries, shows that sedation is not necessary for a successful colonoscopy. “The vast majority of patients are able to be done unsedated,” Dr. Dominitz said. “They are able to do complete colonoscopy in the vast majority of patients, and most patients don't have a great degree of pain.”
Avoidance of pain and anxiety is the primary argument for sedation during colonoscopy. The arguments against include cost, of course, and complication risk. “Most of the complications that come from endoscopy are related to the sedation,” said Dr. Dominitz.
Much research has been done on the subject, comparing complete sedation with propofol, to moderate sedation with benzodiazepines and opiates, to no sedation. Although studies of patient satisfaction are hard to compare, they tend to favor propofol, and the drug is also efficient, with patients being ready for discharge sooner than they would be on a benzodiazepine and opiate, Dr. Dominitz reported.
Propofol is also what most gastroenterology professionals (both physicians and nurses) would choose if they had to undergo colonoscopy, according to a recent study.
But adding cost as a consideration can change decision making. Deepak Agrawal, MD, surveyed more than 800 gastroenterologists and endoscopy nurses for the study, which was published in the Oct. 28, 2013, JAMA Internal Medicine. “After they said, ‘I want propofol,’ I asked them, ‘How much would you be willing to pay out of pocket for it?’,” Dr. Agrawal, an assistant professor of medicine at University of Texas Southwestern in Dallas, told DDW attendees.
The answer was that about 40% of the physicians wouldn't pay anything for propofol and “overall, about 75% of endoscopists and nurses combined said they wouldn't pay more than $200,” Dr. Agrawal reported.
Considering that propofol adds between $150 and $650 to the cost of colonoscopy, this makes it an obvious target for cost reduction, the experts said. Several payers have already refused to cover it for low-risk patients, and more may follow, or restructure payment systems so that gastroenterologists earn less, rather than more, when procedures include anesthesia. “What if in future they take away our anesthesia provider fees?” asked Dr. Agrawal. “Would we still use [anesthesiologists'] services then?”
Reduced reimbursement for anesthesia could be an effect of bundled payments, which are likely on the way for colonoscopy, predicted Joel V. Brill, MD, ACP Member, who updated attendees on the American Gastroenterological Association's work to develop a colonoscopy bundle for colorectal cancer screening and surveillance before payers implement one of their own choosing.
The proposed bundle, which excludes therapeutic procedures, would begin with preprocedural care, such as initial evaluation of the patient, and include all professional and facility costs of colonoscopy, as well as after-care, such as bleeding complications and the need for any immediate re-dos.
“Should you be compensated each time the patient has to be brought back because the prep is poor, or you didn't perform a complete examination, or should you be incented to do it right the first time?” asked Dr. Brill, medical director of FAIR Health, Inc., and an assistant clinical professor of medicine at the University of Arizona in Phoenix. Patients should bear some responsibility, too, experts at the meeting noted, since their compliance with bowel prep affects the results, but how that could work is still uncertain.
This new world of colon cancer screening, with different technologies and payment methodologies, may sound intimidating, but the core task will remain the same, the experts reassured DDW attendees. “Colonoscopy is still going to be around,” said Dr. Lieberman. “It's going to be very important that we do it well, that we demonstrate that we do it well.”