Jargon Muddies Radiology Reports, Including About Cancers

Nick Mulcahy

August 02, 2017

The practice of radiology uses some of the most sophisticated and expensive technology in all of medicine, such as MRI and computed tomography (CT).

But when it comes to summarizing what they see, radiologists typically use a set of clunky catchphrases, such as "diagnostic for," "consistent with," and "concerning for" — modifiers that even they don't precisely comprehend.

In the case of suspected and actual cancers, this antiquated lexicon can wreak havoc, especially when patients access their reports on increasingly available health system portals, experts told Medscape Medical News.

But change is afoot as practitioners from the community, academia, and professional societies take a critical look at jargon, seek greater clarity, and propose and use more standardized reporting.

A lot of work needs to be done, reform-minded radiologists seem to agree.

Jennifer Kemp, MD, a diagnostic radiologist at Diversified Radiology in Denver, Colorado, is one of those professionals. In April, she opened her Twitter account and tweeted, "Old habits are hard to break…wish me luck for my best attempt at jargon-free day!"

Dr Kemp's high-spirited language consciousness is an outgrowth of a cancer trauma in her family. In 2004, her husband was diagnosed with stage III rectal cancer at the unlikely age of 37 years.

The experience partially reset her professional mission.

"I was absolutely shocked that the terms that I had grown so accustomed to were phrases that my husband found confusing and sometimes alarming," she told Medscape Medical News.

"I would try to explain his radiology reports to him and use terms like 'can't rule out' or 'indeterminate,' and he was only confused and dumbfounded," she added.

The language of radiology now deeply matters to Dr Kemp, whose subspecialty is body imaging (CT, ultrasonography, and MRI of the chest, abdomen, and pelvis), which includes a lot of cancer imaging. "This is not to say that I have entirely stopped using these terms; sometimes I just can't find better words to use. But I think twice before I use radiology jargon."

I think twice before I use radiology jargon. Dr Jennifer Kemp

In Denver, Dr Kemp leads a focus group in her practice "to try to improve our report clarity group-wide."

Dr Kemp's efforts represent a relatively recent phenomenon. "When I trained in radiology, there was absolutely no discussion of ambiguous phrases and no sense that there was any need to change our verbiage," said Dr Kemp, who finished a fellowship only 18 years ago, in 1999.

Andrew J. Gunn, MD, from the University of Alabama Birmingham, shares Dr Kemp's concerns. He is one of an emerging number of academicians who has studied the "imprecise" and "varied" language that is ubiquitous in radiology reports, especially the terms that are used in a summarizing fashion. And he has come to a stunning conclusion about his profession.

"The reality is — I don't even know if radiologists know what most of these terms mean," Dr Gunn told Medscape Medical News.

The imprecision — and related confusion — exists in all kinds of radiology reports, including those about possible metastatic cancer, he said.

He points to his 2016 study that evaluated physicians' understanding of 10 common "modifying" terms in radiology reports with regard to metastatic cancer.

The study, for which Dr Gunn was lead author, was a survey of 59 radiologists and 100 primary care physicians at Massachusetts General Hospital (MGH), which he undertook during his training there (J Am Coll Radiol. 2013;10:122-127).

Both groups of doctors were asked to consider hypothetical radiology reports and interpret the statistical likelihood (0% to 25%, 26% to 50%, and so on) of the presence of metastatic disease based on the terms used in the report.

The modifying terms were "diagnostic for," "represents," "likely represents," "probably," "consistent with," "compatible with," "concerning for," "suspicious for," "may represent," and "cannot exclude."

The respondents from both groups were veteran doctors at MGH, with an average of more than 15 years of practice.

The study found that phrase with the most discrepancy between the two physician groups was "consistent with" metastatic disease (but there was discrepancy for all phrases). About three quarters of the radiologists assigned this phrase a 100% or a 75% to 99% statistical likelihood of true metastatic disease. But less than half of the primary care physicians assigned those high likelihoods to the phrase "consistent with" metastatic disease.

This one slice of data from the study also revealed something highly notable about the radiologists.

Dr Gunn explained that some phrasing has been believed to be nearly universally understood among radiologists. "To say something is 'consistent with' metastatic disease — most radiologists would say that is a slam-dunk diagnosis," he said. But the study indicated that, in fact, about one quarter of the radiologists did not see a slam dunk when they read the phrase.

Most radiologists would say that is a slam-dunk diagnosis. Dr Andrew J. Gunn

Furthermore, the study showed that most primary care doctors do not widely think of the phrase that way, and yet another study revealed that patients "definitely don't" understand the phrase that way, summarized Dr Gunn, citing a subsequent survey with MGH patients (Am J Roentgenol. 2014;203:1034-1039).

"There is a disconnect between what we are saying and people are understanding — and that's a problem," said Dr Gunn, who added that the consequences include increased risks for error and confusion.

Origin Story?

Where did this nebulous language come from, and how do radiologists learn it? Dr Gunn says that it is part of the descriptive tradition in radiology.

"You learn it very early on in your training — how you hear your attending say things is how you are going to say things," he said.

But every radiology trainee also learns the lack of uniform usage.

"When I was at Mass General, I would notice on what seemed to be the same finding on an ultrasound or a CT, one of my attendings would describe it one way and another attending would describe it another way," he commented.

Dr Gunn also said the nomenclature is not discussed in medical school and is not purposely taught in residency. "You pick it up from who you are around," he said.

"Historically, radiology reports have just been prose on a blank sheet of paper. People would dictate as they looked [at images] and read," said Dr Gunn.

Some progress has been made to update the "free text" basis of reporting, said William Thorwarth, MD, chief executive officer of the American College of Radiology (ACR). But progress toward greater standardization has been at a "glacial" speed. "We need to accelerate the pace," he told Medscape Medical News.

We need to accelerate the pace. Dr William Thorwarth

A balance needs to be struck in reporting so that radiologists can "give their opinion" in free text but deliver details in a "more structured" format, he said.

Both Dr Thorwarth and Dr Gunn pointed to the Breast Imaging Reporting and Data System (BI-RADS) as a major success in radiology reporting uniformity and standardization. The system, which was initiated in the late 1980s, has been organized and developed by radiologists and the ACR. An important component of BI-RADS is the lexicon, which is a dictionary of descriptors of specific imaging features.

The system also uses a 0 to 5 final assessment scale that Dr Gunn roughly translates as follows: 0:  "I don't know what it is — we need a follow-up exam"; 1:  normal; 2: benign; 3: follow-up in 6 months; 4: may be cancer, needs biopsy; 5: absolutely cancer (>95% likelihood).

The ACR has since developed other similar systems, and most have implications for assessing cancers, including C-RADS (colonography) Li-RADS (liver), Lung-RADS (lung), and TI-RADs (thyroid).

However, Dr Thorwarth pointed out that only BI-RADS is used for every breast image in the United States because it is mandated by federal law. The other reporting systems are elective.

"We are attempting to decrease variation [in reporting] as much as possible," he said.

Additionally, Dr Thorwarth said the Radiology Society of North America (RSNA) has developed dozens of templates for radiology reporting for a wide range of uses, from cardiac radiology to vascular imaging. RSNA says its reporting initiative is "improving radiology reporting practices by building IT standards and a library of clear and consistent report templates."

Obfuscating Jargon

The Journal of the American College of Radiology (JACR) now features a regular column, "Speaking of Language," wherein Jenny Hoang, MBBS, a neuroradiologist at Duke University in Durham, North Carolina, tackles radiology's communication problems.

In October 2015, she offered this fictional radiology report: "Impression: Right lower lobe consolidation most likely represents pneumonia. Cannot exclude malignancy."

Dr Hoang was blunt: "Radiologists should avoid using the phrase 'cannot exclude.'" She explained that this is a "catchphrase" and that if there is an alternative diagnosis, radiologists should say so directly. She offered an improved report: “Findings most likely represent pneumonia. Given the known risk factors for malignancy, a repeat study in 6 weeks would help identify an underlying mass.”

"Radiologists' skills are most valuable when they are used to make diagnoses, not exclude them," she concluded.

Radiologists' skills are most valuable when they are used to make diagnoses. Dr Jenny Hoang

In March 2016, Dr Hoang observed that the management of cancer is often dictated by any difference between two radiology studies over time. Therefore, if there is no change, "just say so," she advised.

The commonly used word "stable" should also be avoided as it can "baffle" patients, she said.

Other terms, such as "similar," "no significant change," and "no substantial change," are more nebulous and rob patients of the clear and certain statement that their conditions have not advanced, she also said.

Dr Thorwarth wondered whether the practice of radiology may eventually move toward having two reports: one for referring physicians and one for patients.

"The precision of medical language is not always compatible with patients' vocabularies and comprehension," he said.

Nevertheless, patients are increasingly privy to radiology reports. Dr Gunn pointed out that the "explosion of patient portals" in the last 5 to 10 years has meant that patients often see their radiology report before the referring physician does. "There are plenty of stories of patients showing up in their doctor's offices in tears," he commented.

"Patients say that the number one problem with radiology reporting is that the language is overly technical or confusing," said Dr Gunn.

The number one problem with radiology reporting is that the language is overly technical or confusing. Dr Andrew J. Gunn

In oncology, the most common radiology request is to determine whether there is treatment response, he further commented.

In a reform-minded mode, Dr Gunn said that "line number one in such a report should be what is the response to treatment" (eg, no change, shrinking/eliminating tumors, stopping progression). The information should be simple, clear, and obvious, he said.

That sounds a lot like what Duke's Dr Hoang preaches.

In June 2015, in her JACR column, Dr Hoang offered a piece of advice that might be described as her reporting creed and a blueprint for change: "Radiology reports should be direct, unambiguous, and factual. Medical terms should be used only when necessary. Obfuscating jargon should be eliminated."

The commentators have disclosed no relevant financial relationships.

Follow Medscape senior journalist Nick Mulcahy on Twitter: @MulcahyNick

For more from Medscape Oncology, follow us on Twitter: @MedscapeOnc


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.