The idea behind screening asymptomatic patients for potentially serious medical conditions has obvious attractions. What if we could detect oncoming illnesses in patients before signs or symptoms develop? We could presumably then intervene in time to save lives, costs, and suffering—this is the principle behind proactive screening, and it certainly sounds great. The trouble is, almost none of medicine’s efforts to achieve such a result have been effective in saving lives.
Last week, the American Cancer Society changed its recommendations on colon cancer screening, lowering its recommended age for first screening from 50 to 45. The impetus for this change, which so far is not being adapted by other major expert bodies including the United States Preventive Services Task Force and the American College of Gastroenterology, is an increase in the rates of colorectal cancer, or CRC, in adults 45 to 50.
The increase in CRC in adults aged 45 to 50 is simultaneously large and minuscule. The large way of looking at things is that it has risen almost 50 percent in the past two decades. That sounds and is impressive, but through an epidemiologic lens, the absolute risk remains small: The rates of CRC in this age group rose from around 1 in 6,000 up to 1 in 3,200. That’s still a fraction of the rates in older adults, and it’s also worth noting that it’s been mainly driven by increased incidence among whites. (The rates of CRC in blacks, while higher than whites, has remained steady—in fact, for the first time, the rates of CRC in whites and blacks younger than 50 appear to be the same. The American College of Gastroenterology recommends that black Americans get screened starting at age 45 already.)
Currently, we simply do not know whether screening asymptomatic patients in younger age groups will change mortality rates. We can theorize that lowering the age of screening has the potential to catch more cases earlier, but it’s critical to note that the decision to expand screening on a national scale to include adults 45 to 49 is based on zero direct clinical evidence. And the problem with that is that we tend to overestimate the benefits of medical interventions like screenings without recognizing or tracking the potential harms.
This is part of why experts are split on the expansion of recommendations. Some, like Harvard’s Sapna Syngal, a gastroenterologist at Brigham and Women’s Hospital, think it’s warranted given what we know about screening in older age groups and the rising incidence. But others feel that given the unclear overall benefits of screening asymptomatic patients for a variety of conditions and the complete lack of evidence in this age group for CRC, this new strategy will merely expand an already shaky paradigm. In other words, the story of colon cancer screening is still being written, and how you feel about the new expansion likely depends on how you feel about screening more broadly.
To most people, colon cancer screening is synonymous with colonoscopy. In reality, less invasive stool-based laboratory tests and special CT scans are almost as effective, far less burdensome and expensive, and unlike colonoscopy, they incite less fear in patients, meaning they have much higher participation rates. That’s why the Canadians and the Europeans have spurned colonoscopy as the mainstay of screening for CRC, instead opting for stool tests that check for the presence of microscopic blood or suspicious DNA, or for flexible sigmoidoscopy (a far less invasive visual test that assesses just the last part of the colon). So why did Americans get hooked on colonoscopy? The short answer is that colonoscopy is slightly more sensitive and, of course, far more expensive.
From the late 1960s to the mid-1980s, colonoscopy—in which a tube-shaped camera is inserted into the rectum allowing doctors to visualize the entire colon in search of cancerous polyps—went from near science fiction to commonplace. Today, the major guidelines in the U.S. recommend the procedure for adults starting at age 50. In 2000, Katie Couric famously had her first colonoscopy performed live on the Today show, a stunt that was apparently sufficient to spark a short-lived increase in the number of screening colonoscopies performed in the United States. But none of these guidelines answer some important questions: Do we have high-quality evidence that colonoscopy actually saves lives? How effective are other less invasive and less expensive tests?
The short answers are that we do not have high-quality evidence that colonoscopy saves lives overall, though indeed some studies suggest it may, and that stool-based tests are both far less expensive and almost as effective.
Syngal thinks that the reason colorectal cancer screening procedure differs so much between countries is that “[colonoscopy] is cost effective, but it is still expensive. They can’t afford it, but we can,” Syngal told me.
If you’re confused, you’re not alone. How can something be both cost-effective and too expensive? Either colonoscopy reduces cancer-related medical costs downstream or it does not. But this is hard to assess, because it’s hard to know how, exactly, to value a human life. If, for example, it costs a medical system $1 billion worth of colonoscopies to save one life, it might not be worth going bankrupt over. But if it only costs the system $50,000 to save a life, it would be unethical not to perform these tests. It may be that the Canadians and Europeans simply draw the line at a different point than we do—we may be willing to spend $3 million to $7 million on testing to save one life (a rough estimate that reflects our general approach here), while other nations might draw the line at, say, $2 million.
The trouble is that we do not know the true costs. In Britain, one study found that 1,250 patients needed to be screened for colon cancer to save one single life. At $3,000 per procedure, that means it would cost $3.75 million in colonoscopies per life saved. But in that study, the margin of error was gargantuan, and it might be that as few as 690 patients must be screened to save one life, or it might be as many as 9,090 must be screened. That means that using colonoscopy as the screen for colon cancer could cost anywhere from $2.07 million per life spared to $27.3 million. Cruel as it sounds, no system could tolerate the latter and stay solvent. “The bottom line is whether additional cost of screening patients at an earlier age is warranted,” says Andrew T. Chan, a gastroenterologist at Massachusetts General Hospital and Harvard Medical School, and senior author on one of the most important studies of CRC screening that suggests a long-term survival benefit stemming from the procedure.
Costs aside, though, there remain differing views on whether more asymptomatic people should be screened at all. “Pushing more screening also has potential detrimental effects, such as more overdiagnosis,” John Ioannidis, a professor of medicine, health policy, and statistics at Stanford University told me via email. False positives and overtreatment from screening for diseases like prostate and breast cancer are dangerous enough—and occur with such frequency—that any decreases in death due to those cancer screens are likely offset by these risks, risks that are vastly underreported in studies. Colonoscopies, for example, can cause major bleeding or life-threatening intestinal perforation, which can lead to sepsis and death. While rare, these risks must be weighed against the admittedly small number of patients who benefit from CRC screening, especially in younger cohorts.
Nevertheless, all screening is not created equal and Ioannidis believes CRC to be one of the few success stories. Still, he is skeptical about the new ACS guideline. “The data for moving the start age to 45 instead of 50 are weak. It may not be wrong, but my interpretation is that the extra benefit is questionable and may not exist at all,” he says. Even proponents of colonoscopy, including Andrew Chan, agree with this. “We don’t know the long-term benefits of screening in this age range and that’s where we have a deficit and knowledge gap,” Chan adds. Recognizing this, even the ACS describes its own new recommendation as “qualified.” And unlike its unabashed enthusiasm for colonoscopy in patients older than 50, the guideline suggests presenting younger patients with several options, including the stool-based tests favored elsewhere.
Meanwhile, randomized studies that compare colonoscopy to stool testing alone and even colonoscopy versus no screening at all are finally underway. These trials will bring us much closer to definitive answers on whether these regimens are effective and economically sustainable. In the meantime, the expansion of colon cancer screening is likely to continue.
The question is what should people in the 45 to 49 age range do right now? Bearing in mind that the new ACS guideline is based on no direct clinical evidence at all, and that other prominent expert societies have not adapted this expanded approach, it may be most prudent for younger patients without risk factors (such as a strong family history of CRC) to wait for new evidence to emerge before getting screened for CRC. Caution is especially warranted for younger patients considering colonoscopy as the means of testing. Despite the increased rate of CRC in people younger than 50, the absolute rate remains so minuscule that even rare complications from colonoscopy may turn out to harm more people than could ever have been helped. We simply don’t know yet. Meanwhile, stool-based tests, and even CT scans, may be viable and safe alternatives. But considering the flimsy ground upon which the entire case for expanded CRC rests, for now at least, it’s reasonable to sit this one out.
The opinions expressed in this article are solely those of the author and do not reflect the views and opinions of Brigham and Women’s Hospital.