Even best medical reporting infected with “make do” bias

Press Criticism | By Craig Gurian |

While the association was stronger for cancers on the left side than on the right, the results “indicate a clinically meaningful reduction in risk of colorectal cancer [CRC] death with colonoscopy throughout the colon and rectum.”

Remapping Debate interviewed Dr. Baxter, who confirmed that, “if your definition of ‘gold standard’ is highest accuracy, then it’s going to be colonoscopy” as compared to other procedures.

A large U.S. population-based case-control study showed that colonoscopy was associated with a material reduction in the odds of colon cancer mortality, “indicating a substantial protective effect.”

We also spoke with Dr. Alfred Neugut, a professor of medicine and epidemiology at Columbia University Medical Center. Neugut confirmed that his position had been, as of two or three years ago, that it was uncertain the extent to which colonoscopy is better than sigmoidoscopy. The latter procedure, as explained to Remapping Debate by Dr. David Johnson, a former president of the American College of Gastroenterology, “cover[s] only about a third of the colon, 40 percent, and we know that…nearly 40 to 45 percent of polyps in cancers are above the reach of [that] short scope.”

Neugut, however, has written more recently that the Baxter study was “game-changing” and demonstrated that the colonoscopy is associated in a reduction in CRC mortality “clearly superior to that of sigmoidoscopy.”

In short, those Americans who heeded advice to have colonoscopies were, in the main, doing the best thing for their health. It’s hard to imagine that this wouldn’t have been reported on if the atmosphere were not suffused with the “Cadillac care” narrative.

(I should point aout that it may be the case that some patients without suspicious symptoms or a family history of colon cancer are being prescribed follow-up colonoscopies too frequently. A study cited by Rosenthal did find almost 25 percent of this subset of the study cohort had a follow-up colonoscopy in fewer than seven years, as opposed to the generally recognized recommendation of a 10-year follow-up after a negative result with no intervening symptoms. But the article in The Times failed to mention that the study was limited to patients at least 66 years of age. Individuals aged 50 to 65, widely believed to be under-screened for colon cancer, were not, as the cited paper itself made clear, part of the study.)


Apples and oranges

Rosenthal was not wrong to indicate that there are those (including the U.S. Preventive Services Task Force) that describe screening methods other than colonoscopy as acceptable. That position should not, however, been accepted at face value.

Take, for example, the “fecal occult blood test” (FOBT). Performed annually, it is included as one of those acceptable screening methods. (If an FOBT is positive, of course, the next step is a colonoscopy.)

“If your definition of ‘gold standard’ is highest accuracy, then it’s going to be colonoscopy” as compared to other procedures, said Dr. Baxter.

Dr. Jason A. Dominitz is the National Program Director for Gastroenterology at the Department of Veterans Affairs. In an interview with Remapping Debate, he said that FOBT could “potentially detect the important polyps before it’s too late,” noting that “many polyps never become cancer.”

Doesn’t this leave out an important fact about what an FOBT can do?

Dr. Brian Jacobson, an associate professor of medicine at Boston University School of Medicine and chair of the Health and Public Policy Committee of the American Society for Gastrointestinal Endoscopy, provided an answer, pointing out to Remapping Debate that polyps “tend not to bleed unless they’re very large,” meaning that FOBT is “not great as a polyp detection test.”  FOBT, in other words, is “really looking to detect cancer when it’s still in an early and, therefore, hopefully curable form. So it’s not so much of a preventive test as [it is] an early detection test,” Jacobson said.

Now Dominitz was not himself arguing that FOBT is a cancer prevention procedure, and, in fact, emphasized to us that he was not “anti-colonoscopy” and that he agreed completely with the proposition that an important advantage of colonoscopy is that a colonoscopy can remove polyps before they become a problem.

Nevertheless, in casual discourse, the nuance of what tasks different screening methods can perform is lost in the fog of purported “equivalence,” even though most people, if asked, would not see cancer prevention and cancer detection as the same thing. They’re not, and the difference deserved more than Rosenthal’s characterization of colonoscopy as merely “intuitively” appealing. Likewise, her discussion of the “anointment” of colonoscopy in 2000 by the American College of Gastroenterology as a “the preferred strategy” for prevention falsely pictured Americans as foolishly acting like sheep in following what appears to have been a sound recommendation.


Missing the importance of who performs the procedure

The article in the Times failed to apprise readers that some of the studies that had questioned the relative superiority of colonoscopy had looked at patient populations where a high proportion of colonoscopies were not performed by gastroenterologists, something that may have caused the potential of colonoscopy to be understated.

Multiple experts we spoke to agreed that there is great variation in the skill level of those who perform colonoscopy, both between specialties and within a specialty.

Baxter said that “colonoscopy is really quite a complex skill and [there is] quite a variation in providers.” Her study found that the association with reduction in cancer was “significantly stronger” for colonoscopies performed by a gastroenterologist as compared with a surgeon, primary care physician, or other doctor.

Send a letter to the editor