The NordICC study has provoked intense debate in the medical community on the role of colonoscopy screening for colorectal cancer (CRC). But perhaps more than anything, the debate has exposed a misunderstanding of what the NordICC trial was really about. The debate has also revealed that medical opinions have become largely polarized and subjective, rather than nuanced and objective.
In this column, I will attempt to clarify some misconceptions surrounding the interpretation of the NordICC trial circulating on social media.
The NordICC trial was a pragmatic randomized controlled trial (RCT) involving asymptomatic people 55-64 years of age drawn from population registries in Poland, Norway, Sweden, and the Netherlands between 2009 and 2014. Participants were randomized in a 1:2 ratio to either receive an invitation for a single screening colonoscopy or to not receive any invitation. Of the 28,220 who were invited, 11,843 (42%) accepted the invitation and underwent screening.
At 10-year follow-up, the risk for CRC in the overall intention-to-screen population — also known as intention-to-treat in treatment trials — was 0.98% vs 1.2% in the control arm, revealing a significant lower relative risk of 18%. The risk for CRC-related death was 0.28% in the intention-to-screen population vs 0.31% in the control arm — this difference was not statistically significant. The risk for all-cause death was similar in the two arms.
And finally, in the per-protocol analysis, which only included the 11,843 patients who received a colonoscopy, the risk for CRC-related death was 0.15% vs 0.30% — a 50% lower relative risk; however, the risk for all-cause death was not provided.
This trial has led to a storm of controversy. Here's a breakdown of some key concerns and how we can look at the trial results objectively.
One major argument circulating among experts is that the study title — "Effect of Colonoscopy Screening on Risks of Colorectal Cancer and Related Death" — is misleading.
The argument is that the trial wasn't designed to evaluate the effectiveness of colonoscopy screening. Instead, the RCT was set up to explore whether an invitation to undergo one colonoscopy effectively lowers the risk of getting and dying of CRC.
Given the focus of the trial, some have asserted that people should not conclude that colonoscopy screening is ineffective; but rather that the invitation to undergo a single colonoscopy is not very effective — and certainly less effective than we might have expected.
I agree with this take. And I think a more accurate conclusion to the trial would be that an invitation to receive a single colonoscopy in Poland, Norway, Sweden, and the Netherlands reduced the risk for colorectal cancer but did not decrease CRC-specific or all-cause mortality compared with people who did not receive such an invitation.
However, for brevity, in medicine we usually phrase these as the effect of interventions. Ultimately, all interventions tested in RCTs are invitations to participate. Nobody can force people to undergo a medical intervention.
But the discomfort with the title reflects a deeper issue with the trial. Less than half of the people (only 42%) invited to colonoscopy actually had one. This is not only a critical limitation of the study, but it is an important finding in and of itself: Population-level uptake of colonoscopy as a screening modality is low.
One argument is that the benefit would have been apparent if everyone who had been invited underwent a colonoscopy. But that's fiction. In reality, no population-level screening program will have 100% uptake rate, especially for something as unpleasant, and potentially costly, as a colonoscopy.
However, the 42% rate is lower than observed in the US population. According to data from the Centers for Disease Control and Prevention, in 2018 about 60% the eligible population received a screening colonoscopy.
So, a better question would be: If the uptake rate were 60%, would the results of the NordICC trial change and by how much? I would say probably not much. Even when looking at only those who got a colonoscopy in the trial, the risk for CRC-specific death was only 0.15% lower.
Another potential point of confusion: The public debate around the trial has conflated individual, patient-level decision-making with population-level health policy. The NordICC trial addresses the efficacy of a population-level colonoscopy screening decision. Put another way: The question these researchers were trying to answer is whether we should include population-level colonoscopy screening as part of our health policy programs. That's why randomization at the invitation level makes sense.
This confusion between individual and population-level decisions stems from the difference between an intention-to-treat, or in this case an intention-to-screen, analysis and a per-protocol analysis. The intention-to-screen analysis — which the NordICC trial used — preserves the benefit of randomization. RCTs try to see the effects of making decision A vs making decision B. Everything that may or may not happen after this decision remains a part of the study, which should ultimately tell us whether that decision improves outcomes.
A per-protocol analysis — which only compares patients who completed the originally allocated treatment with those who did not — is cheating because it doesn't study the outcomes of decision-making; it only examines the outcomes of the modality, in this case colonoscopy.
Imagine this scenario: You have a very effective cancer drug that is designed for four cycles of treatment but has a 10% chance of a fatal adverse event within one cycle. You cannot run an RCT of the drug against another agent and remove patients who couldn't complete four cycles of the investigational drug, analyzing only the ones left after four cycles. Similarly, the NordICC trial, which tested colonoscopy screening programs, needed to include everyone who was randomized to an invitation for colonoscopy; it could not exclude those who decided not to undergo colonoscopy.
That is why I believe this was a well thought-out, well-conducted trial. It allows us to scrutinize whether population-level colonoscopy screening invitations work and answer an important policy question: Should countries invest in population-level colonoscopy screening?
The answer, based on these results, is that inviting asymptomatic, average risk people to screening colonoscopies has a limited impact on the risk of dying of colorectal cancer or any cause.
Importantly, though, the NordICC trial doesn't tell us that all colorectal screenings are ineffective. It also doesn't tell us whether you, the individual, should get a colonoscopy screening.
It simply means that colonoscopy screening programs appear to be relatively ineffective in reducing mortality at the population level. For countries considering investing in population-wide colonoscopy screening programs, the NordICC trial essentially indicates that money is probably better spent elsewhere on programs with a stronger mortality benefit.
This concept may be unusual, especially in countries like the United States that have no population-wide national screening programs. Screening in the United States is more a function of individual patients and their interactions with their doctors or insurance companies.
Another point to highlight here is the difference between absolute and relative risks. It is important to look at both and decide whether those risk differences are worth the intervention. In NordICC, there was no significant difference in CRC-specific or all-cause mortality in the intention-to-screen population. When limited to the per-protocol analysis, the risk reduction for CRC-specific mortality was only 0.15% in absolute terms, but 50% — a reduction from 0.3% to 0.15% — in relative terms. Whether you interpret this as is a substantial risk reduction is an individual judgment call.
In the wake of the NordICC trial, the American Society for Gastrointestinal Endoscopy and others have argued that colonoscopy should remain the gold standard for CRC screening. That is a question that deserves to be revisited on the basis of evolving evidence, not necessarily assumed to be true.
Finally, it's important to clarify the difference between colonoscopy as a screening test and a diagnostic test. NordICC is about screening colonoscopy in asymptomatic people with average risk, which is different from a diagnostic test in people with symptoms, such as blood in the stool.
If you have any concerning symptoms, you should undergo an evaluation without delay.
With screening tests, emotions are heavy. We all know someone who was saved by screening or someone whose cancer wasn't diagnosed early because they skipped being screened. For us, these instances weigh far more heavily than aggregate data in studies. Many of us have already undergone a screening colonoscopy, and many of us are contemplating it in next few years.
This is a personal decision, and a decision that needs to be guided by data and reason, not emotion.
Bishal Gyawali, MD, PhD, is an associate professor in the Departments of Oncology and Public Health Sciences and a scientist in the Division of Cancer Care and Epidemiology at Queen's University in Kingston, Ontario, Canada, and is also affiliated faculty at the Program on Regulation, Therapeutics, and Law in the Department of Medicine at Brigham and Women's Hospital in Boston. His clinical and research interests revolve around cancer policy, global oncology, evidence-based oncology, financial toxicities of cancer treatment, clinical trial methods, and supportive care. He tweets at @oncology_bg.
Lead Image: Bishal Gyawali, MD, PhD
Medscape Oncology © 2022 WebMD, LLC
Any views expressed above are the author's own and do not necessarily reflect the views of WebMD or Medscape.
Cite this: A Controversial Trial: Exposing Misunderstandings of NordICC - Medscape - Oct 14, 2022.