Virtual Colonoscopy Can Be Hazardous to Your Health

The US Preventive Services Task Force (USPSTF) recommends that everyone aged 50-75 be screened for colon cancer with any one of three options: colonoscopy every 10 years, flexible sigmoidoscopy every 5 years, or fecal occult blood testing (FOBT) every year. Conventional colonoscopy is considered the “gold standard” since it allows for direct detection and biopsy of early cancers and removal of precancerous polyps. It involves passing a long colonoscope via the rectum through the full length of the colon and is also known as optical or visual colonoscopy. A newer and less invasive alternative, virtual colonoscopy or CT colonography, is being promoted by some as the test of choice. Others disagree. One area of controversy is that CTs frequently find “incidentalomas” that require further investigation. An article in the journal Radiology highlights this problem, describing “the clinical drama that follows screening or diagnostic tests.”

Virtual colonoscopy, or CT colonography, has some advantages:

  • No sedation required.
  • Avoids the discomfort and the small risk of bowel perforation that accompany colonoscope insertion.
  • Takes less time.
  • Visualizes right colon (not well visualized in 10% of optical colonoscopies).
  • It is as effective as optical colonoscopy in finding advanced cancers, but results in far fewer polypectomies.
  • May identify abnormalities outside the colon.

It also has some disadvantages:

You will notice that identification of abnormalities outside the colon is listed as both an advantage and a disadvantage. Assessments of risk/benefit ratio differ. In a 2005 study of 500 patients published in Radiology, CT colonoscopy found “a substantial number” of clinically significant findings, including aneurysms and a renal carcinoma; but it did not establish whether finding them improved outcomes. 28% of the significant findings had already been diagnosed prior to colonoscopy; and the patient with the renal carcinoma declined treatment, since he already had metastatic colon cancer. In a 2009 study of 143 patients, 98% of patients had at least one incidental finding outside the colon. 24% of these required further evaluation, with 73 imaging studies, 30 lab studies, 44 clinical visits, 6 medical procedures, and 44 return visits over a mean period of 38 months. The cost was estimated at $248 per patient. They did not assess whether these evaluations improved eventual patient outcomes.

Writing in the journal Radiology, Dr. William Casarella, a radiologist at Emory University School of Medicine, tells the story of his own virtual colonoscopy. His colon was normal; but the CT also showed areas outside the colon, revealing a kidney lesion, a 2 cm. mass in the liver, and multiple non-calcified nodules in the bases of both lungs. These findings led to the following tests and interventions:

  • A contrast-enhanced CT of the abdomen (with more radiation) showed that the kidney lesion was a benign cyst, but the liver mass was not.
  • A high-resolution lung CT (more radiation) confirmed lung nodules.
  • A CT-guided (more radiation) liver biopsy (more risks) showed necrotic tissue, no definitive findings.
  • A PET scan (more radiation) was negative.
  • Video-aided thoracoscopy
  • 3 wedge resections of the right lung (requiring the lung to be collapsed), resulting in a diagnosis of histoplasmosis (benign, no treatment indicated).

The lung surgery led to:

  • 5 hours in the recovery room before awakening from general anesthesia.
  • Chest tube, subclavian central venous catheter, nasal oxygen catheter, epidural catheter, arterial catheter, subcutaneous heparin injections, prophylactic antibiotics, IV narcotics (each with associated risks).
  • Excruciating pain requiring narcotics for 2 weeks.
  • 4 weeks of disability.
  • Persistent chest pain due to surgical interruption of intercostal nerves.

The total cost was over $50,000.

This is an atypical example, but it underlines the potential dangers of too much information. Sometimes ignorance is bliss. Dr. Casarella would have been better off had he not been screened for colon cancer at all. I support the USPSTF recommendations, but it is important not to over-rate the value of these tests. The public perception is “get this test: it will save your life.” The reality is more complicated. Screening tests can have false positive and false negative results. They don’t always provide clear black and white answers, and they don’t save as many lives as the public tends to think. While statistically benefiting the population, they may be hazardous to the health of some individuals. We need to keep in mind that even the best screening tests have a downside.

Posted in: Cancer, Diagnostic tests & procedures

Leave a Comment (19) ↓

19 thoughts on “Virtual Colonoscopy Can Be Hazardous to Your Health

  1. woo-fu says:

    We need to keep in mind that even the best screening tests have a downside.

    So true!

    Even with traditional tests, like the regular colonoscopy, problems are on the rise. Some doctors are rushing through the procedure. I actually overheard a nurse discussing how one gastro doc was the fastest in town. “He can do three in the time it takes most doctors to do one!” she said. The procedures are done at the office, but at that speed I have to wonder how thorough the doctor is. Plus there’s an increased risk of perforation.

  2. windriven says:

    Very thought provoking post Dr. Hall. One can read your post and still conclude something along the lines of ‘better discomforted than dead.’ Or one can read it and wonder how much basic, high quality medical care could have been delivered for the $50k spent on making one guy miserable.

    As you probably know, the US spends about $4500 per capita per annum on health care while other developed countries average somewhere around $2300. Unfortunately, that differential does not buy the US proportionately longer life expectancy. Cuba spends less than $500 per capita and achieves nearly the same life expectancy as the US (1). It may be that our high rate of spending on health care buys improved quality of life but I have not seen credible data that supports that conjecture.

    Health care is a finite resource. The US spends will spend nearly 20% of GDP on health care by the end of this decade (2). Our political leadership has carefully avoided an adult dialogue with the American people about health care spending and what it buys and what it costs (not just in dollars but in opportunity cost).

    If the status quo endures we can expect higher spending on health care without commensurate extensions of life expectancy. And we can certainly expect more spending on quackery. Homeopathic snake oil is cheaper than, say, positron emission tomography. If it makes the patient and their family happy and it costs less, our vigilant protectors in DC will certainly give it their warm embrace.


  3. cervantes says:

    Colonoscopy, nevertheless, has one of the best benefit/risk profiles of the screening tests, largely because it can a) actually prevent cancer and b) there’s very little problem of overdiagnosis because as far as we know there’s no such thing as an indolent colon cancer. Colonoscopy can also definitively diagnose and characterize diverticula and other conditions, which may be a small bonus in some cases.

    These problems you mention are confined to CT scanning, which does find plenty of incidentalomas while having less of the upside of colonoscopy.

    That said, colonoscopy is definitely a pain in the ass, no joke, and it costs more than other screening procedures.

  4. Emma B says:

    Of course, that little anecdote wouldn’t be nearly as affecting if the lung resections had showed cancer that could be successfully treated, rather than a benign disease.

    Three months ago, my husband (age 35, father of three young children) went to the doctor with some vague-but-nonspecific symptoms. A chest x-ray showed a shadow that might have been an enlarged spleen or an image artifact, and the doctor encouraged him to get an ultrasound to check it out further. That ultrasound showed that his spleen was fine, but that he also had a 5cm mass on his kidney. One CT scan and one partial nephrectomy later, he had a diagnosis of renal cell carcinoma, stage 3a. If we had left it alone and not done the ultrasound until he developed clearer symptoms, we would very likely have been dealing with an advanced stage 3 or a stage 4 cancer, with a very poor prognosis.

    Renal cell carcinoma has historically been a deadly cancer because it doesn’t usually cause symptoms until it’s well advanced. However, it is increasingly being detected in its earlier stages as an incidental finding on CT scans done for other reasons, and outcomes are improving.

    When we tell just-so stories about unnecessary testing, let’s not forget that there are patients like my husband whose lives ARE saved by aggressive pursuit of incidental findings.

  5. maeris says:

    I’ve had several arguments with my fiance about this. He believes the future of medicine holds Star Trek-like scanning devices, which instantly identify any problem non-invasively. I’ve argued with him numerous times that knowing everything is dangerous, for exactly the reasons discussed in this article. More research needs to be done to identify the benign from the problematic, to prevent these unnecessary surgeries. If, in the future, we could indeed easily identify all problems in the system (Star Trek style) we could do this sort of research much more easily. However, until all screening becomes fast, cheap, non-invasive and easy, accurate identification of false positives will remain elusive.

    That being said, like many findings in science, polarization of the issue (we should screen vs. we should not screen) is a disservice to everyone. We should advocate screening as well as a cautious interpretation of the findings.

  6. Harriet Hall says:

    @Emma B,

    “let’s not forget that there are patients like my husband whose lives ARE saved by aggressive pursuit of incidental findings.”

    No. Let’s not. But let’s not read too much into anecdotes like yours. It is quite possible that your husband might have developed some symptom in the next few days that would have prompted testing and resulted in the same outcome. There is no way to know whether such incidental findings really save lives except by doing controlled studies. If the outcomes from renal carcinoma are improving, there is no evidence that screening CT scans are responsible. If incidentally finding kidney cancers on CT screening were proven to save lives, that would mean we should be systematically screening the entire population with CT for early detection of kidney cancers. I don’t think anyone would seriously advocate that.

  7. SloFox says:

    The question regarding the potential benefits v. risks of incidental findings is captured in large part in the whole body imaging fad. I could be wrong but it seems that it came and went. When I stopped following it there had been no demonstrable benefit to whole body imaging. I know I’m being lazy but does anyone have any different data at hand?

  8. windriven says:


    Dr. Hall pointed out: “[CT colonography] may miss smaller polyps and even some larger ones: CT colonography failed to detect a lesion measuring 10 mm or more in diameter in 10% of patients and missed 15% of advanced cancers measuring over 6 mm.”

    There seem to be at least two huge variables in ‘whole body imaging': the resolution of the technology and the attention and skill of the interpreting radiologist.

    One can still get a whole body CT scan in Taiwan for about U$D 2000. I know at least one person who does it with some regularity. But like you I haven’t seen any data that suggests that it is useful in extending life or improving quality of life. Given the exposure to radiation, the missed positives, the cost of the procedure and interpretation by a qualified radiologist, the morbidity and pure scaredshitlessness attending the biopsying and analyzing false positives, it is hard to imagine data sufficiently compelling to recommend it for routine screening.

  9. mxh says:

    @Dr. Hall, I agree with the main point of your post, but you seem to be guilty of the same thing that you’re accusing Emma B of. You’re using an anecdote (that, as you say, is an atypical case) to promote your point. The danger of CT screening isn’t incidentalomas, it’s how the physician approaches the incidentaloma. The same is the case for breast cancer screening, PSA, and many other screening methods.

  10. Harriet Hall says:

    How the physician approaches the incidentaloma? Are you seriously thinking that any doctor would ignore these findings? Even if he wanted to, the patient would insist on finding out what those spots meant, and the doctor would be afraid of a malpractice lawsuit if he failed to investigate and missed a serious diagnosis.

    Guilty of the same thing? Not exactly. I provided an illustration to make a point, and I said it was not typical. I said we shouldn’t read too much into Emma’s anecdote, and of course we shouldn’t read too much into Dr. Casarella’s case, either. It only serves as a reminder that screening tests can have unexpected adverse consequences. I offered a published case report that was a clear example of harm and Emma’s personal anecdote could not be interpreted as proving that a life was saved.

  11. woo-fu says:

    Anecdotes aside, the major point that I picked up from Dr. Hall’s post is that some put way too much faith in what can be gleaned from diagnostic imaging, without adequately being prepared for or informed of the risk and cost. An image isn’t necessarily the big picture. Luckily, most of my doctors are in agreement. I have to have imaging done more frequently than most to monitor medical conditions which have already been diagnosed.

    From my perspective, I don’t expect doctors to be wizards or images to be crystal balls, nor would I want them to be. However, others do place an unwarranted amount of faith in medicine, modern or alternative. At least, in my experience, most doctors do not encourage such blind faith and are frank about drawbacks. On the other hand, blind faith is often at the very heart of CAM marketing strategies and practices.

    This is a good reminder to apply critical thinking regardless of how esteemed a practice may be.

    @EmmaB Very glad your husband had such a positive outcome!

    @Windriven I think it’s been a year or two since I stopped seeing adds in the local paper for mobile imaging clinics. They would announce their scheduled stops, and for a few hundred dollars they promised all kinds of scans and tests. No doctor I spoke with ever recommended them.

  12. windriven says:


    “[I]t’s been a year or two since I stopped seeing adds in the local paper for mobile imaging clinics…”

    Great point! I never used them so their disappearance didn’t register with me. I wonder if the cause was market conditions or regulatory intervention?

  13. Ed Whitney says:

    Later today the results of CT screening for lung cancer will be announced. This involved annual screening for high-risk populations (30 plus pack-year smoking history) for three years, and there were reductions in both lung cancer and in all-cause mortality. What to do with all those nodules will of course be a big issue, and it will be interesting to see how the news media handle the findings. The USPSTF will look at it soon, and their rating will be interesting to see (Noodle-Nose Vinnie is laying 2:1 odds on a B rating). Later of course the Obamacare death panels will have to weigh in, but this is going to be a real test of how to apply and interpret data in the age of evidence-based medicine. Not to mention market-based medicine.

  14. BillyJoe says:

    “I wonder if the cause was market conditions or regulatory intervention?”

    In Australia, clinics appeared offering exercise ECG for anyone who wanted them. They were all bulk billed under Medicare meaning the patient paid nothing. After an outcry by cardilologists, medicare stopped paying for these and the clinics disappeared. Good result.

  15. JMB says:

    Virtual colonoscopy was originally promoted not because it was a better test than colonoscopy, but because it was a better test than a barium enema. Compliance of the population to the screening recommendation of colonoscopy for colorectal cancer is poor. Some people were willing to have the barium enema but not the colonoscope. The CT virtual colonoscope is not perfect, but the barium enema is even less perfect.

    The promotion of whole body CT screening scans stopped because several state legislatures passed laws or state boards of health issued guidelines against them. After anecdotal reports of major complications from biopsy of common incidentalomas like adrenal or lung nodules, CT screening soon was restricted by politics. The American College of Radiology released this position paper in 2002 which discouraged whole body CT screening of an unselected population.

    That may have helped those state legislatures restrict whole body scans. Of course, now we have all those people with ultrasound machines offering screening for stroke risk and heart/ peripheral vascular disease.

    Above a certain level of spending, there is very little relationship between life expectancy and per capita healthcare spending. Life expectancy in the US is currently on the rise because of a significant drop in smoking that began about 30 years ago. However, the rise in obesity is offsetting some gains. Accidents, smoking, and obesity are the big factors in average life expectancy. That life expectancy gap will shrink regardless of whether we have a government agency tell us how to spend your money, or if we let people decide themselves. Cuba has also been known to fudge on it’s healthcare data. Wait and see how long Chavez lives after having his cancer surgery in Cuba. If you have diabetes or cancer, your life expectancy is longer in the US than in many of those countries with better average life expectancy.

    As far as the USPSTF decision on lung cancer screening, I wouldn’t hold my breath. There have been clinical trials reported for many years showing the benefit of CT screening for lung cancer in patients with 30+ pack year smoking history. However, there has been only one large scale multicenter RCT necessary for the USPSTF to give it an A or B recommendation.
    The cost would be prohibitive. But rather than talk about the prohibitive cost, they’ll probably bring up issues of anxiety and risks of false positives, overdiagnosis, and overtreatment.

    Now there is an interesting sidelight to the USPSTF, virtual colonoscopy, screening topic. The USPSTF recommendations for colorectal cancer screening (2008) preceded the recommendations for breast cancer screening (2009). The decision analysis in 2008 was the first use of the efficient frontier analysis in making the recommendations. The efficient frontier analysis of virtual colonoscopy explicitly included the cost of the exam in the analysis. GE, a major manufacturer of CT scanners, objected to the use of the cost data in that fashion. However the rule was never changed. Subsequently, when the USPSTF used efficient frontier analysis in the decision about mammography (Nov 2009), they left out specific cost data, but used the number of exams as a proxy.

    The following are links useful to assess the argument in the preceding paragraph,

    The GE pdf is unchanged. The gov web pages have been modified from their original versions in Nov 2009.

  16. mxh says:

    Dr Hall. My point wasn’t that doctors shouldn’t do anything about incidentalomas, it’s more that pretty much every other screening test has the same problem and, therefore, incidental findings shouldn’t be a reason to not perform a screening test. What’s more important is how the doctor prepares the patient for the results of the test and how they inform them of the tests risks (including the possibility of discovering incidentalomas). Incidentalomas or other questionable findings are not unique to virtual colonoscopies and saying that “X test can be hazardous to your health” is misleading to patients.

    Perhaps a solution could be that patients are informed of the possibility of incidental findings and it’s their choice whether they want questionable findings investigated (of course, clear-cut findings are a different matter). This is a common approach used with research subjects in medical imaging (and I’m not aware of any lawsuits from incidental findings that later proved to be life-threatening).

  17. Harriet Hall says:


    I agree that incidental findings are not a reason not to perform a screening test. They are only one of many considerations in deciding when to test and what test to use. I would argue that more incidentalomas are found with virtual CT than with regular colonoscopy, and that should be factored into the individual patient’s decision.

    My point is that patients (and some doctors) often get the idea that screening tests are an unadulterated good that uniformly saves lives, and they should be told that the reality is more complicated and that there are pros and cons.

  18. woo-fu says:

    My point is that patients (and some doctors) often get the idea that screening tests are an unadulterated good that uniformly saves lives, and they should be told that the reality is more complicated and that there are pros and cons.

    So very true! Medical practice is messy. Tests aren’t perfect. Labs aren’t perfect. And every patient presents a unique set of issues and reactions.

    Speaking of which, if a patient is having continual mid-gut problems and neither a colonoscopy or endoscopy can reach the problem area, would CT be advised in that case? Or are there other techniques, besides exploratory surgery, that are better?

    That some patients still view doctors as gods presents major problems when the gods they’ve created can’t deliver. It creates an almost adversarial relationship.

    I personally prefer a doctor who isn’t afraid to say, “I don’t know.” And I’m happy to report that all of my doctors are very proactive when it comes to discussing risks of tests & procedures. And they aren’t shy to discuss other medical options or to refer me to a specialist when needed. These doctors earn my trust. On the other hand, there are doctors with their own god complex, and I’m trying to stay far away from them.

    This post is very timely for me. Someone close to me is facing the choices you discuss. I think she’ll appreciate this link. Thanks again!

  19. mxh says:

    Dr. Hall, agreed.

    Woo-fu. I totally agree with you on the importance of conveying uncertainty in patients. I’d feel much better with a doctor who says “I don’t know, but I’ll work hard to find out.”

Comments are closed.