The Dangerous Incidentaloma

In medical jargon, an “incidentaloma” is something inside of you that we found by accident. It’s usually used in reference to a radiology procedure like a Cat Scan (computed tomography or CT) or MRI, but it can apply to any accidentally incidentally found thing. From our previous review of statistics and what constitutes “normal” we know that with most blood tests about 1 in 20 will be “normally abnormal,” in that the test comes out abnormal but the person is actually just one of the normal people at one or the other end of a bell curve.  In radiology it’s a bit harder to know how many “abnormal findings” show up in NORMAL people. When one does show up it’s an “Incidentaloma,” which is a play on the usual “oma” ending attached to various cancers such as lymphoma, sarcoma, carcinoma etc. It’s kind of cute at first but the results can kill you.

A few years ago I started to see patients arrive in my office with a report from “Lifeline Screening.” After some questioning I found out that “Lifeline Screening” was a for-profit company that goes around to various places, often churches or community centers, and does this great community service (or so they say!).  For just a mere few hundred dollars they will do a bunch of tests on you that you don’t need (but apparently want because you forked over the cash) and, about half the time, find SOMETHING.

Of course they can’t advise you on what to do about that something except “go to your doctor.” Now from a business standpoint I shouldn’t mind.  After all, they are generating more visits to my office. Unfortunately (for my bottom line) I’m an ethical doctor and take seriously my job of “primum non nocere” (Latin for “First do no harm”).  It made me mad as a hornet, however, to have to clean up a mess that I didn’t make. In addition, by referring these “incidentalomas” to me, I now must assume all the risk (with regards to malpractice etc.).  If I don’t “get to the bottom” of it… ARRRGH!  How do I “cause no harm” if I’m obligated to do MORE tests?  Usually I would sit down, review the data with the patients, go through many of the things I’m trying to put forward in this book, and reassure them that the only bad thing happening was that their wallet was now several hundred dollars lighter.

For doing this consultation – which takes a good long while if done with care – I got paid peanuts. Remember doctors rarely get paid well to talk, think or counsel; the get paid to DO. It would have been SO much easier to just order the follow up study, blood test, whatever, schedule a follow up appointment and move on to my next patient. CHA CHING! Of course every test I do has its own associated false positive rate and now I’m potentially CAUSING MORE HARM.  ARRRGH again!

The most frustrating occasions occurred when I simply could not reassure the patient well enough and was basically forced (by my patient) to order more tests. This happened, unfortunately, often.  Some people, once they know SOMETHING is not normal; that it might be CANCER, simply MUST do something about it. To me diagnosing an incidentaloma is one of the worst harms we doctors do. It’s basically thievery; we’ve stolen away whatever time you would have had in BLISSFUL IGNORANCE.  Blissful Ignorance should really be more valued than it is.  Although “early detection” is the gospel for those who would sell you all these tests, the truth is that starting treatment “early” for many diseases doesn’t really change any outcome except to shorten your period of blissful ignorance.  By that I mean you eventually die at the same age but your “morbidity” is higher.

Morbidity is usually paired together in the phrase “morbidity and mortality.” In fact, most institutions have monthly “M & M” rounds where they discuss (in private) the “mea culpas” of the month and try to figure out how they could have been prevented.  Morbidity is a VERY important thing to study and is closely related to “Quality of Life.” Much to researchers dismay it is very hard to quantify.  Various quality of life indices have been developed but it’s so subjective that telling any one individual patient how treatment will affect his quality of life is difficult if not impossible.

Let me just give you some numbers for which we have fairly good data: they are called adrenal incidentalomas.

The adrenal glands are a pair of very important organs that sit on top of the kidneys. They make various hormones like adrenaline (epinephrine). Rarely, they develop significant tumors that produce excess hormones and cause some serous illness.  Sometimes a cancer from somewhere else in the body spreads to the adrenal glands. Frequently (as in about 85% of the time), however, they have INSIGNIFICANT tumors that don’t make anything, don’t cause any trouble, don’t kill you or hurt you, they are just there.  Think of it like a little mole you’ve had for years that is NOT cancer. Just a mole. Some people even call them “beauty marks.”

The problem occurs when we find one of these when we’re doing a study for another reason and it just happens to be in the area of the adrenal glands. The radiologist, who doesn’t even know you at all, is obligated to report the finding. You end up back in my office with a report that tells me you don’t have any problems with your gall bladder but “by the way” we did notice this “thing” in the adrenal.  ARRGH!  Now what do we do?
Fortunately this is one of the rare versions of an incidental finding that has actually been studied so we can hang our hat on a few statistics. Here they are:
About 4.4 % of CT scans done for other reasons will find an adrenal incidentaloma.  Of these, 85% will be benign (don’t cause ANY harm) and 15 % will be “real” problems.  Many, if not most, of the 15% will have some type of symptom or finding that would have lead an experienced doctor to figure something was wrong anyway, but I digress.

After an adrenal incidentaloma is found by a CT scan, patients get, at a minimum, a follow-up appointment with their doctors.  Now the doctor is in a pickle.  Let’s say he does a complete and thorough exam, history taking etc. but still doesn’t see anything that could explain an adrenal problem.  He will probably still have to order a few more tests.  Some are blood tests, some are urine tests where pee is collected for 24 hours (see endnote for humorous story about pee) etc.  In any case, it’s going to cost the patient some money and probably some body fluids, not to mention time.  This happens 4.4% of the time when we do a “high resolution” CT scan.  Most of the studies that get around a 4% incidence of incidentaloma are when the CT scan is being done for some other medical reason.  In other words something was wrong (at least subjectively) to start with.  When WELL patients go get CT scans the incidence of incidentalomas will probably be a lot higher like the 6-12% range.  Autopsy studies generally find about that amount so as the CT Scanners get better and better this will become an even bigger problem.

To me, if you have to see the doctor again, get tests done, worry etc., then “harm” has been caused.  Let’s calculate the number needed to harm just from ordering a CT scan.

Of the 4.4% of Incidentalomas, 85% will be benign

0.044 x 0.85 = 0.0374

So 3.74% of people will have a benign incidentaloma on CT scan.

Remember to get Number needed to HARM we put 1 over that number

1/0.374 = 27

So, not really TOO bad.  There’s a 26 out of 27 chance the test (the CT scan) won’t end up hurting you.


Now what about the number needed to treat (or, in this case, accurately diagnose a problem)?

4.4% x 15% = 0.044 x 0.15 = .0066

One over that number yields NNT 1/0.0066 = 151


So you can see that you’ve got to find a LOT of these guys to actually help out ONE person.

Even if we just ignore the 96% or so that did NOT have an incidentaloma the picture isn’t good.

Number Needed to Treat = 1/0.15 = 6.7

Number Needed to Harm  = 1/0.85 = 1.17

So you’re MUCH more likely to have a bunch of tests for nothing than we are to find something wrong with you.  All because you got a CT scan for some OTHER reason!

Consider what one of the experts says:

“If the results of hormonal testing are normal and the imaging features are consistent with benign disease, I would recommend repeating the imaging studies at 6, 12, and 24 months and repeating the hormonal evaluation yearly for 4 years, even though there are no data from large, long-term studies to support these recommendations.”[1]

Now I’m really picking on this test because we have some data for it.  I realize the CT’s were “probably” being ordered for perfectly reasonable causes and researchers are going to cry foul with my use of Number Needed to Treat and Number Needed to Harm in this fashion.  I consider it perfectly reasonable, however, to ask your doctor “what are the possible side effects of getting this CT scan?”  I doubt many would say “well there’s a 1 in 27 chance we will have to do a bunch more tests on you for the next 4 years for nothing!”   (Not to mention the CAT scan might get you  – see “Cat Scans can Kill You” page on this blog.)


We just don’t really think like that.  We are trained to DO something and even when we screw you over, MOST patients will end up thinking we did the right thing.  It’s dumbfounding to me, really, but I’ve seen it time and time again.  At the end of the 4 years when the patient is finally discharged (at least from concern about this malady) he says “Thank goodness they caught that and it was nothing.  I really feel better now.”  ARRGGHH!


As you can see, I’m very worried that if everybody starts getting full body scans….it will be a mess.  Please don’t get one.  Pretty Please!



[1]          Volume 356:601-610 February 8, 2007Number 6 The New England Journal of Medicine. The Incidentally Discovered Adrenal Mass

                William F. Young, Jr., M.D.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s