Friday, August 3, 2012

Vague radiology reports: Can anything be done?

Possible mild peri-appendiceal inflammation. So states a CT scan reading I received the other night.

Here’s another. The findings are concerning for appendicitis.

And another. The appendix is dilated and mildly thick walled with suggestion of mild surrounding inflammatory change, although there is air within the lumen. There is an appendicolith at the base of the cecum. The findings are suspicious for mild acute appendicitis. Clinical and laboratory correlation are recommended. [Digression: You can't diagnose appendicitis with lab tests.]

Even radiologists are questioning the way their colleagues dictate reports. From an editorial in the journal Applied Radiology: The report might say “there appears to be a nodule in the right lower lobe.” If you’re wrong, and there isn’t a nodule in the right lower lobe, you’re covered because you never actually said there was a nodule, only that there appeared to be a nodule, and we all know that nothing is ever as it appears.

Some Stanford researchers looked at radiology reports and found that the most frequent terms used to modify conclusions were probable, consistent with, consider, likely, suggestive, no definite evidence, suspicious, cannot exclude, not likely, maybe and possible. They surveyed radiologists and clinicians and found wide variations in what these words meant to each physician regarding the specifics of whether a hypothetical lesion should be biopsied.

An editorial in the journal Radiology scorned the use of the often-seen phrase if clinically indicated.

I could go on.

Do you know of similar examples of radiologic ambiguity?

Is there a solution to this problem?

A version of this post appeared on Sermo yesterday with interesting comments. About 2/3 of those who voted said there is no solution to the problem. Some commenters said the radiologists are just describing what they see.

35 comments:

Anonymous said...

As the surgeon you need to read it yourself BEFORE looking at the radiology report then only use the report to make sure you didn't overlook something. Radiologists are NOT clinicians.

Skeptical Scalpel said...

Thanks for the comment. You didn't have to tell me radiologists are not clinicians. I was aware of that. I read all my own imaging to the extent possible. I also discuss any issues in person with a radiologist. I often get a second radiologist's opinion too. Unfortunately, I can't do in person discussions or second opinions at night when the radiologist is in Australia.

Anonymous said...

I have had the opportunity to talk to overseas radiologist at night with difficult cases and have found them to be very helpful on occasion. The techs can hook you up and tell you how to get in touch with them(and tell you if they are any good). It can be fun to have a brief moment of talking to someone in a more respectful and less hurried culture.

Skeptical Scalpel said...

Yes. I've talked to them too. It sometimes takes a while to get the right one on the phone and you usually do not know them at all. There are at least 40 radiologists in the outsourced group we use.

Óli said...

Having only worked in radiology for the summer I'm in no way an expert, but there is a reason radiology responses are sometimes vague. Radiology isn't just a machine you put patients into and get a diagnosis out of. Clinical correlation is always important and that's the reason it's important to give a good history when requesting a scan.

A mild inflammatory change (dirty fat which is very hard to report objectively or measure) with a mild thickening of the appendix wall is probably appendicitis if it has the correct clinical history. But in a jaundiced patient with RUQ pain this finding would probably be of no significance.

Another example is a fracture line on plain xray of a wrist. If there is no pain and no trauma, it probably isn't a fracture, but the radiologist has no way of examining the patient himself so he'll have to say just that.

Radiologists try to be objective and measure everything that they can, but there are limits. Imaging is very dynamic, movement in the scanner, body fat and peristaltic movement may have an effect on the item measured.

Last point, it varies between radiologists how objective they try to be. Some would write "appears to be a nodule at the base of the lung". I would never write that, I'd decide on my own whether the (small, questionable) nodule is something of significance or not and then say something like "small nodule at the base of the lung of uncertain etiology".

UtilityKnife said...

The biggest issue I hear among my colleagues relating to radiology reports is that the radiologist isn't vague but too directive. Ovarian torsion is a good example. This is a clinical diagnosis and no amount of radiologic findings are going to proscribe precisely what to do in a given patient. In this sense being vague may be appropriate since the radiologist cannot know the context from which to estimate the significance of the radiographic findings. I'd be content with 'just tell me what you see.'

Dr Skeptic said...

Thanks, and I agree that this is a problem.
One idea is that radiologists report imaging studies without any history provided (blind reporting), so as not to bias their reporting.
A colleague recently referred a patient for a hip MRI, looking for FAO (functional acetabular impingement). The radiologist phoned my colleague and said that he really wasn't sure there was much on the scans, but if he wanted, he would write in the report that FAI 'could not be ruled out', or was 'suspicious' or whatever.
Out in private practice, radiologists are running a business and need to keep their referring surgeons happy. Blind reporting would solve that particular problem.

pilocarpine said...

I was taught in medical school that investigations are there mainly to support the diagnosis that already been made based on sound clinical evaluation of history and clinical examination. Therefore,radiological reports will be vague if clinical evaluation has no bearing of any directions.

Skeptical Scalpel said...

Thanks for the comments. I've often wondered why radiologists say they want clinical information too. One, they really aren't good clinicians. Two, what difference does the history really make? Just tell me what you see.

Some radiologists have told me that they put explicit instruction like "consider MRI" because they think some clinicians need help to decide what to do next.

Interesting comment about the radiologist wanting to not preclude the clinician from the diagnosis of impingement. I had not heard of that before but it is a good point.

titobman said...

I have to disagree, radiologists can be very good clinicians. I don´t think that our work (i´m a radiologist) is just describing what we see but instead according to the sex, age, clinical background, signs and symptoms we can offer an accurate diagnosis. Maybe those things aren´t that important for an appendicitis but for other more complex pathology it is. And other thing on giving clinical information and the suspicious pathology, it´s very important for doing a correct protocol for the patient and adjust the radiation dose!, it´s not the same doing a ct to rule out appendicitis, mesenteric ischemia or pancreatitis.
And as everything in medicine, there is a spectrum in diseases. Really advanced and obvious appendicitis to report on one side and minimal, and maybe subjective, inflammatory changes that one can´t totally assure it´s appendicitis on the other.

Anonymous said...

Well I have a very fresh example from two weeks ago.

The house officer (Intern you say in the US ?) called me up and was upset. She had just got a referral for a CT/urography back from the senior radiologist. He refused to perform the examination on a patient who had presented with Gross Hematuria after finding out that I (that is me "Anonymous") a couple of days previously had diagnosed a urinary bladder cancer.

His argument was: "Why should I do this, you have already found the reason of the Gross Hematuria"

It ended with me calling this Senior Radiologist, and he was all upset over the House officer who "dared to demand from him anything"

Between breaths, I explained to him that she was ABSOLUTELY right. He burst out in long explanations of how many years he had been working and that he was a personal friend of this-and-that famous Urologist...blablabla..

Fine I said, in a calm fashion, and continued: If you want to take 100 % personal responsibility for the patient NOT having a Panurothelial transitional cell carcinoma also affecting either ureter and/or the renal Pelvis, please give it in writing...

Needless to say, we got our CT-Urography ASAP.

To make it clear: We - the clinicians are the experts in clinical diagnosis, they - the Radiologists - are the experts in Radiology.

End of Story

Skeptical Scalpel said...

Interesting story. This proves the point that talking to each other leads to understanding.

Anonymous said...

Exactly, and it should be a two-way communication. This is especially important when a junior doctor actually is doing the very right thing. She shouldn´t be shut down by a senior doctor from a service speciality, just because he or she is having a bad day, a chronic ulcer or any other weird reason :)

Anonymous said...

Reading the radiology reports one becomes aware of the carefully drafted prose used by radiologists and, one is aware of the fact that radiologists are trained and indoctrinated to report their “impressions” based on the defensive medicine dictum. Thus, the reports are often vague and miss the intended purpose: to provide an accurate diagnosis and/or assist in diagnosing.

I would like to suggest that the applications of the methodology of evidence-based radiology (EBR) promise a profound influence to improve the validity of the radiological “impression” and thereby diminish the nebulous prose that, at times, we experience whilst reading a radiology report.

The foundation of EBR provides the scientific infrastructure to ascertain validity (Internal validity, external validity, precision, et cetera), and has been introduced as a new approach to the practice of radiology.

The practice of radiology includes both diagnostic and interventional modalities; hence the fundamental concepts of Evidence-based Clinical Practice are pertinent. The Centre for Evidence-Based Radiology in Ireland provides a didactic material tailored to radiologist interested in learning the principles of EBR as they apply to diagnostic and therapeutic imaging. For those radiologists interested in advance training in the field of EBR, fellowship training is also available in radiological informatics and radiological imaging.

When clinicians order radiology investigations, they are asking now days: (1) what is the post-test probability of disease? Or (2) Will the interpretation of the test(s) results or “impression” alters the probability of a disease so that the clinician is confident in starting the best therapeutic intervention or (3) conversely, based on the radiologist impression, the clinician can exclude a suspected disease or pathology. If the latter is true, the clinician will seek further consultation from the radiologist to select the most appropriate test.

After reading the radiology report, the ordering physician can combine the diagnostic accuracy or end-points provided by the radiologist, e.g., sensitivity, specificity, likelihood ratios, with the accuracy derived from the physical findings, as highlighted in Evidence Based-Physical-Diagnosis.

Thus, by mastering the foundations of EBR, specifically as they relate to diagnostic or interventional radiology the reports will be devoid of nebulous jargon and will express instead, valid measures that aid with confidence in the clinical decision making process.

In terms of medical malpractice avoidance, if the radiology reports are narrated with diagnostic accuracy, the diagnostic report(s) will stand in its merits as evidence, and accepted clinical practice, thus standard of care.

Anonymous said...

I agree with the most recent post. I have to call you out on a couple of things Skeptical;

On one hand, you want the radiologist to simply tell you what he/she sees without any context and provide an accurate diagnosis. The less information they receive about patient history (which should be collected by the technologist or written on the prescription), the more vague the report is likely be. You want a diagnosis from the radiologist without any background information, and what little is provided is vague itself.

On the other hand, you are expecting them to put their ass on the line with this little amount of information and you complain when they ask for more. Do you not understand that appearances on any single radiological examination can be similar for mulitple diagnoses? Sometimes multiple imaging studies are needed to collect enough information for a positive diagnosis. There is no one-stop exam that can evaluate every pathology. When the radiologist asks you for clinical information, it is usually for two reasons; To back up the suspected diagnosis or to provide a lot of evidence that the prelim was incorrect. If they suggest another type of examination, it is because they aren't sure of a diagnosis based on the amount of info they collected from the original exam... it's to help you out. Believe it or not, they are on your team!

It is far too common these says for physicians of any specialization to avoid looking at the patient in their entirety. That is, most only focus on their specialty without asking any questions outside their area of expertise that might be relevant to a diagnosis. Whether because of ego, defensive medicine, plain laziness, or even stupidity patients are forced to put the pieces together themselves based on little tidbits of information gathered by because of this apathy from each specialist... with a little help from google. The medical field is outraged at the attempts of lay people to find information on their own, but as a health care provider myself, I suggest it is because we are not giving them the information they are looking for (and google is free). What else are they supposed to do?

I applaud any radiologist who has an open mind to consider pertinent information outside of what is presented on the monitor before them, and any physician who extends beyond their comfort zone to consider more than just their specialty.

I'm surprised to see you stereotype an entire group of physicians as "not good clinicians". It shows a lot of ignorance on your behalf, and a narrow-minded view of the radiology field. Where would your career be without radiologists?

Skeptical Scalpel said...

I appreciate your taking the time to comment.

I am fortunate to have worked with mostly excellent radiologists over the years. Except for IR, I have rarely if ever seen a radiologist even touch a patient, let alone examine one. They may know clinical medicine but I don't think they are really clinicians.

I should have put this thought in the blog. I think many of us are a little jealous of the fact that radiologists can say what they think, but in most cases we (e.g., surgeons) have to make the difficult decisions like whether to operate on a patient.

"Possible mild peri-appendiceal inflammation" is not helpful in the decision-making process. If that is truly what was seen, then so be it. I had to decide what to do at 3 in the morning. The Radiologist went on to read the next CT.

Yes, I know *I* chose to be a surgeon and not a radiologist.

Anonymous said...

Here I have to reply to Anon of his/her posting [August 11, 2012 11:03 AM]- I think Sceptical Scalpel is really on the right track: Radiologists are (in general) really good Radiologists. But not Clincians in the way we define the Clinician. The Clinician has a constant influx of data, the Patient story combined with labtests, combined with physical examination, combined with Radiology-evidence, and not to forget: most often a process going on for days and sometimes weeks (in complicated cases), when you have to judge, weigh and integrate all incoming data, past data and the continuos flow of changes in all different aspects. The Clinician has the total integrated picture of the patient - this is not the case for the (EXCELLENT) Radiologist (and we truely respect all of you and cherish your very helpul accomplishments). To make it easier: Is the Radiologist also a good Anaestesiologist ? Well No - and that goes for the majority of us dealing with surgical matters too. Is the Radiologist also a good labdoctor/Virologist/Psychiatrist/ whatever-speciality-we can-think-of too ? Of course not - AND that is ther very same for us Surgeons or Urologists too. I have had the great pleasure to work with some really highly qualified and seasoned Uro-Radiologists, in different hospitals, for the last 20 years - and they have been really great, efficient, updated, intelligent, clever and very professional in their work,serviceminded and extremelly accurate -but are they good clinical Urologists ? My reply is: No.

Skeptical Scalpel said...

Thanks for the support. I think you made my point better than I did.

Anonymous said...

Please forgive me if this is a redundant post; I have not read all the posts here.

I always believed a surgeon should look at ALL x-rays. . .even a trivial pre-op "normal" CXR.

Having said that, if it is 2 am, the ER doc calls me for an appy, I get-up, look at the CT, act on it, come to surgery, remove the appendix and dictate my findings in my pre-op H and P of what the CT or radiograph says, then I, not the radiologist, should be paid for the reading. It does NOT count when I get a phone call, in the recovery room, by a radiologist, who says, "Doc...did you see that young lady last night...she has appendicitis."

"Sure...did...operation was done 2 hours ago."

Trim waste in medical spending right there. Radiologists are "consultants" too...why "consult" them and ask them to READ everything if you don't need them to read "everything." I don't need them to read everything...if I'm not sure...I'll "consult" them...otherwise, I'll order the CT, act on it.

No...I actually LOVE radiologists...they're great where I work. They teach me stuff all the time, and I show them specimens and operative videos. It is a team. If they screw up, I call them, and ask them to dictate an addendum...and they will. If I screw up, they call me and say, "Hey...you missed one."

True teamwork. Great people...but I don't always need them...and they don't always need me.

Anonymous said...

Oh..one more thought. Because they are "consultants" and "doctors" when we ask for a consult, like a clinician, they should come down to the ER (or the Ward), get a stethoscope, perform a complete consultation, exam, look at labs, prior films, talk to family, and then recommend a radiographic intervention/test. This crap of "Every patient with an abdominal CT MUST have oral contrast or I won't read it because it increases MY liability" should warrent a visit to the ED, by the radiologist, to put the NG tube in, get puked on by the distended tender, abdomen of the obstructed patient, and explain why you're going to pump in a half-gallon of gastrograffin while they're puking because you need a "better study". Oh..and they should make rounds and follow-up their "consults", too.

Too damn much Prima Donna crap. Consultants? Yes you are...act like one. And you're a doctor, too.

Skeptical Scalpel said...

A radiologist with a stethoscope? Please send me a photo of that when and if it ever occurs. I think you have a better chance of getting a picture of the Loch Ness monster.

Anonymous said...

Just a message to Skeptical Scalpel - If ever you required any form of imaging for yourself or your family member for a complaint, I doubt very much you would be relying on your own interpretation skills of it as a surgeon. If you want to interpret images - get board certified first!

Skeptical Scalpel said...

Yes, I would prefer a timely, unambiguous reading by a board- certified radiologist who I could go over the images with even at 2 am. Unfortunately, that is not always possible.

Anonymous said...

Why would radiology reports be amended to look like original reports?What happens to the original report? Defensive medicine?

Skeptical Scalpel said...

I'm not sure what your questions are about but I'll try to answer.

Reports can be amended as long as the change is dated and timed. This is automatic with electronic records. The original report is unchanged. The amended text goes in as an addendum at the bottom of the original.

It's not necessarily defensive medicine. The radiologist could have re-reviewed the images or asked a colleague to look at them. Since the original is still there, there is no way to change it to be more defensive.

radiology programs in chicago said...

I was looking for something about radiology programs. Glad to come across this post. Good job.

Anonymous said...

Would it be highly unusual to find 10 changed radiology reports in you records.

Skeptical Scalpel said...

Depends on the time frame, but 10 in one or two years would be very unusual.

Anonymous said...

I think 5 in the same year. Possibly more.

Skeptical Scalpel said...

I'd say that's a significant number of changed reports. You might have a quality issue.

Anonymous said...

No sure what that means?

Anonymous said...

My oppion is radiologists dont really care about what they are looking at, to rushed to report what they
See and have no one to answer to when extremely vague reports are an issue. I dont see doctors
Writing reports and challenging what is written by radiologists or not written. If radiologists have the
Final say then the law needs to be more firm about how films are read.

Skeptical Scalpel said...

Thanks for commeneting. I think most radiologists do care about what they are looking at. I have writtennotes in charts saying I disagreed with a radiologist's report. It happens.

Kat said...

I want to know how a radiologist can say:

FINDINGS: An enhancing 4 x 3 cm enhancing focus posterior to the right
lobe of the thyroid is suggestive of an adenoma (series 10 image 41). A
venous structure is noted to extend inferiorly from the lesion.

Dimensions were given. This is suggestive of an adenoma. So if it's not and adenoma, what is it? Parathyroids are tiny and not easily detectable.

Skeptical Scalpel said...

Kat, most radiologists would not give a definitive diagnosis such as adenoma, which must be made by a pathologist. It could be a carcinoma. To call it an adenoma would be potentially misleading.

You are correct that a normal parathyroid is measure in mm not cm. Even a parathyroid tumor would rarely grow to 4 x 3 cm.

Post a Comment