Colonoscopy is an excellent test for evaluating patients with rectal bleeding, persistent diarrhea, abdominal bloating, and unexplained iron deficiency anemia. Most colorectal cancers are detected because of symptoms,^1,2 and virtually all will necessitate a diagnostic colonoscopy. But most of the estimated 15 million colonoscopies performed in the United States each year are for asymptomatic patients.^3,4 Given polyp detection rates approaching 50% in average-risk populations,⁵ one might start to think that the most common indication for colonoscopy is a prior colonoscopy. High volumes combined with generous remuneration—anywhere from $1,000 to $5,000⁶—has turned colonoscopy into a big business. Colonoscopy profits help explain why gastroenterology practices have become a prime target for private equity.^7,8

Colonoscopy is not a simple screening test; it is a labor-intensive procedure. Some patients require days of preparation. All must consume only clear liquids or a low-residue diet and consume a bowel purgative to clear the colon the day before colonoscopy. Inadequate preparation is common, occurring in approximately 20% to 25% of cases.⁹ Then there is the trip to see a team of clinicians (eg, gastroenterologist, anesthesiology professional, nurse, technician).

There is no evidence that colonoscopy is any more effective than the fecal immunochemical test (FIT). Earlier this year, a randomized controlled trial that included more than 57,000 patients found that FIT was noninferior to colonoscopy regarding risk of colorectal cancer mortality at 10 years.¹⁰ Similar trials are ongoing in Sweden and the United States, and all report higher screening participation rates in the FIT group.¹¹ There is also no evidence that patients prefer colonoscopy to stool testing; if anything, the preference is for noninvasive stool testing.^12–14 Stool DNA testing with FIT (Cologuard) is another screening option recommended in guidelines, although with a caution that this testing has lower specificity and more false-positive results than FIT alone, leading to more unnecessary colonoscopies.¹⁵

So why has colonoscopy become the de facto “gold standard” for colorectal cancer screening in the United States? One reason is good marketing (ie, encouraging the presumption that examining the entire colon to remove the most polyps possible is the best way to lower colorectal cancer mortality rates, although recent data suggest otherwise).¹¹ Another reason is that gastroenterologists have become so good at it, making the procedure easier on patients. Finally, insurance shields most patients from the actual cost of colonoscopy, which is approximately 100 times that of a FIT.¹⁶

There have been long-standing concerns about the overuse of screening colonoscopies. Many Medicare beneficiaries with a negative screening colonoscopy result had a repeat examination within 7 years without a new indication,³ a problem illustrated in the Lown Right Care case scenario in this issue of American Family Physician.¹⁷ Not only do more colonoscopies beget more colonoscopies, they also beget more complications. An estimated 2 to 3 million low-value colonoscopies are performed each year.¹⁸ Although their risk is relatively low for an individual, the large number of procedures produces large numbers of adverse events (ie, approximately 8,000 serious bleeding events and approximately 2,000 bowel perforations).

The overuse of colonoscopy is emblematic of a larger problem. Too many specialists (not only gastroenterologists, but also urologists, dermatologists, and mammographers) have become dependent on revenue from cancer screening, as have the health systems for which they work. Screening is a great way to attract new patients. Although few will have cancer, many will be said to have signs of cancer (requiring further evaluation) or be at elevated risk for cancer (requiring further monitoring). Providing routine services to patients who are well is easier and more lucrative than caring for the sick. Well patients are more easily scheduled during business hours, less likely to experience complications from interventions, and more likely to be well insured.

These incentives discourage a thorough consideration of the trade-offs of cancer screening—although a few may benefit, many more will be harmed by increased worry, aggravation, out-of-pocket costs, and unnecessary interventions and the associated complications.¹⁹ Despite its aggressive promotion, screening is at best a mixed bag and not one of the most important things we do in medicine. Declines in cancer mortality are largely the result of smoking cessation and improved treatment, not screening.²⁰ Although screening is asserted to save lives, there is no evidence that screening an average-risk population helps people live longer (ie, results in lower all-cause mortality). Although screening is claimed to reduce disparities, cancer mortality disparities persist despite similar screening rates.²¹

Preventive interventions are overwhelming primary care.^19,22 What do we want primary care physicians to do: investigate the well or care for the sick? It is time to get back to what truly matters: symptoms.²³ We can help those who are sick and suffering. And we can avoid turning people into patients under the guise of prevention.