Original Research
Health Care Policy and Quality
September 22, 2016

Differences in Perceptions Among Radiologists, Referring Physicians, and Patients Regarding Language for Incidental Findings Reporting


OBJECTIVE. The purpose of this article is to compare radiologists', referring physicians', and patients' interpretations of expressions within radiology reports to describe findings of likely low clinical significance.
SUBJECTS AND METHODS. Surveys were completed by abdominal radiologists (n = 13), physicians referring patients for abdominal CT (n = 59), and outpatients awaiting imaging (n = 51) at a large urban academic medical center. Surveys presented 10 expressions for describing an incidental 5-mm liver lesion and asked respondents to select from a list of choices their perceived likelihood that the lesion represented malignancy. Radiologists and referrers were asked supplemental questions.
RESULTS. Compared with radiologists' concern, referrers' and patients' concerns were higher for four and seven of the 10 expressions. Only the expression “benign cyst” was associated with no concern in all groups; “most likely a cyst” and “too small to characterize” were associated with median levels of concern of 0% for radiologists, > 0% to 1% for referrers, and > 2% to 5% for patients. Expressions containing the phrase “not excluded” had the highest concern in all groups. Referrers' likelihood of ordering follow-up imaging varied widely for the expressions (e.g., “benign cyst,” 2%; “cyst,” 22%; “most likely a cyst,” 46%; “most likely a cyst, although tumor not excluded,” 75%). Overall, the preferred phrase for a 5-mm liver lesion with benign features in normal-risk patients was “cyst” among radiologists and “benign cyst” among referrers. Seventy-six percent of referring physicians thought that radiology reports should indicate whether follow-up imaging is recommended for such lesions.
CONCLUSION. Ambiguity in radiologists' language for incidental low-risk findings may contribute to increased patient anxiety and follow-up testing, warranting greater radiologist attention and potentially new practice or reporting strategies.
Incidental findings on cross-sectional imaging are common [1], but they pose a clinical challenge. Although the exact cause of the finding may not be known with certainty (e.g., a tiny low-attenuation liver lesion on CT that is suggestive of a cyst or hemangioma), such lesions are most often benign and will be of no clinical significance for the patient [1, 2]. Nonetheless, it is standard practice for radiologists to describe such findings within radiologic reports to maintain completeness and caution, given the possibility, even if very low, that the finding may in fact represent a malignancy.
The reporting of a most likely insignificant incidental imaging finding can itself be associated with substantial negative repercussions. The mere mention of the finding in the report can heighten anxiety for the patient [3], who may not fully appreciate the lesion's very low probability of ever causing any harm. Even the referring physician may share similar uncertainty regarding the lesion's potential significance. As a result, incidental findings can result in a cascade effect, with a subsequent stream of additional imaging tests, if not invasive procedures, to become absolutely certain that the finding is benign [1, 2]. These tests can result in substantial costs, inconvenience, radiation exposure, and possible morbidity [4].
Given these concerns, it is critical that radiologists communicate clearly their level of concern when reporting low-risk incidental findings. Nonetheless, past studies have observed variation in how referrers interpret language used by radiologists to convey their level of confidence in a reported diagnosis [5, 6]. Specifically, different referrers may perceive a single expression used by radiologists as corresponding with widely variable levels of certainty, thus providing a source of ambiguity within radiology reports [5, 6]. However, earlier studies did not focus specifically on language used to describe low-probability lesions. This focus may be of particular importance given the frequency of such lesions as well as the high potential for their significance to be misunderstood or taken out of context. Moreover, these studies did not consider how patients perceived such language, which is of increasing importance given a growing tendency for patients to read their own radiology reports [7]. Such insights could be useful for efforts to help referrers and patients alike properly comprehend radiologists' perceived level of concern regarding low-risk incidental lesions.
The aim of this study was to compare radiologists', referring physicians', and patients' interpretations of expressions that may be used within radiology reports to describe findings of likely low clinical significance.

Subjects and Methods

This prospective HIPAA-compliant study was approved by the NYU School of Medicine institutional review board. The study consisted of three separate anonymous surveys of radiologists, referring physicians, and patients at the author's institution, a large urban academic medical center (NYU Langone Medical Center). The decision to complete the survey served as consent for participation in the study. Electronic surveys were conducted using a web-based survey tool.

Survey of Radiologists

Those members of the abdominal imaging section of the institution where this study was conducted who routinely read abdominopelvic CT examinations (n = 14) were solicited to participate in the electronic survey. The survey presented 10 expressions potentially used within radiology reports to describe an incidental 5-mm liver lesion (Table 1). Respondents were asked to indicate for each expression their perceived likelihood that the described lesion represented a malignancy, selecting from among the following choices: 0%, > 0% to 1%, > 1% to 2%, > 2% to 5%, > 5% to 10%, > 10% to 20%, and > 20%. Radiologists were also asked to consider the scenario of a 5-mm sharply marginated homogeneous low-attenuation (less than about 20 HU) lesion in the liver detected on a routine portal venous phase CT performed in a 50-year-old patient with generalized abdominal pain who had no known history of malignancy or hepatic disease. These features were selected to correspond with the “benign imaging features” described by the flow-chart for incidental liver lesions in the American College of Radiology's white paper on managing incidental findings on abdominal CT [1]; the flowchart indicates that a 5-mm lesion with such features does not require any further follow-up regardless of the patient's risk level. The radiologists were asked which of the 10 expressions they thought was optimal for describing such a lesion as well as whether such a lesion requires follow-up imaging.
TABLE 1: Responses by Radiologists, Referring Physicians, and Patients Regarding Perceptions of 10 Potential Expressions to Describe an Incidental 5-mm Liver Lesion
ExpressionMedian Perceived Likelihood of Malignancy (%)Referring Physicians Who Would Order Follow-Up Imaging (%)
Cyst0> 0-1> 0-122
Benign cyst0002
Most likely a cyst> 0-1> 0-1> 1-246
Most likely a cyst, although tumor not excluded> 2-5> 2-5> 5-1075
Tumor not excluded> 10-20> 5-10> 5-1075
Tumor not entirely excluded> 5-10> 5-10> 5-1083
Too small to characterize> 0-11-2> 2-536
Too small to characterize but most likely a cyst> 0-1> 0-1> 2-529
Statistically likely to represent a cyst> 0-1> 0-1> 2-531
Tumor considered unlikely> 0-1> 1-2> 1-225

Survey of Referring Physicians

A departmental database was searched to identify the 1000 referring physicians who had most recently ordered an abdominopelvic CT for an adult patient at our institution. These physicians were then solicited via e-mail to participate in the electronic survey. Of these, 24 individuals were later excluded on the basis of being duplicate entries for the same physician or having incomplete or inaccurate e-mail contact information, leaving a total of 976 referrers for whom an e-mail was successfully sent. Referrers were asked to indicate for the 10 expressions their perceived likelihood that the described lesion represented a malignancy, selecting from among the previously noted list of choices. They were also asked to indicate in a binary fashion whether they would order follow-up imaging in response to each of the 10 expressions. In addition, the referrers were asked which of the 10 expressions they thought was optimal for describing the lesion noted in the previously described scenario. Finally, they were asked to indicate their specialty as well as their level of agreement (1–5 scale, with 5 indicating highest level of agreement) that the radiology report should include a statement explicitly noting whether follow-up imaging is recommended for such a lesion.

Survey of Patients

A survey was conducted of adult outpatients awaiting a cross-sectional imaging examination in our department. The survey could be completed manually, by using printed copies of the survey that were made available to patients within waiting areas within departmental outpatient imaging facilities, or electronically by following a URL indicated within a flier that was also posted in the waiting areas. Patients were asked to indicate their age, sex, highest level of education, and whether they were employed in health care. In addition, patients were asked to indicate for the same 10 previously noted expressions their perceived likelihood that the described lesion represented a malignancy, selecting from among the same choices.


The seven categories regarding perceived likelihood of malignancy were ranked from the lowest (0%) to highest (> 20%) option, and the median option was identified for each of the 10 expressions for each of the three surveyed groups. For the scenario presented to radiologists and referrers at the conclusion of the survey, the most commonly selected option in terms of the preferred expression for describing the presented lesion was identified. Results for remaining survey items in the three groups were summarized descriptively using percentages. Statistical assessment was performed using Excel for Mac (version 15.20, Microsoft).


A total of 13 of 14 radiologists (93%), 59 of 976 referrers (6%), and 51 patients (denominator and response rate indeterminate) completed the survey. Of the referrers, 61% (36/59) were in a medical specialty or subspecialty, 24% (14/59) were in a surgical specialty or subspecialty, and 15% (9/59) were in emergency medicine. The patients had a mean (± SD) age of 54 ± 15 years (median, 57 years; range, 24–88 years); 51% (26/51) were men and 49% (25/51) were women. Patients' highest completed level of education was high school diploma in 16% (8/51), undergraduate degree in 33% (17/51), master's degree in 33% (17/51), doctorate or comparable professional degree in 14% (7/51), and other in 4% (2/41) of cases. Among responding patients, 12% (6/51) indicated that they were employed in health care.
Table 1 provides the median responses by the three groups in terms of the perceived level of likelihood of the described finding representing a malignant tumor for the 10 possible expressions. For the phrase “benign cyst,” the median response was 0% in all three groups. However, for four of the remaining nine phrases, the median responses for referrers corresponded with a higher percentage than the median response for radiologists, and for seven of the remaining nine phrases, the median responses for patients corresponded with a higher percentage than for radiologists. For example, for the phrases “most likely a cyst” and “too small to characterize,” the median responses were 0% for radiologists, > 0% to 1% for referrers, and > 2% to 5% for patients. For all three groups, expressions containing the phrase “tumor not excluded” or “tumor not entirely excluded” were associated with the highest perceived likelihood of malignancy.
Referring physicians' responses regarding whether they would order follow-up imaging for the described lesion varied widely among the expressions. The percentage indicating that they would order follow-up was 2% for “benign cyst,” 25–36% for an additional four of the phrases, 46% for “most likely a cyst,” and 75–83% for expressions containing “tumor not excluded” or “tumor not entirely excluded.”
When asked to select the optimal phrase for describing a 5-mm liver lesion with benign imaging features in a patient without increased risk factors for malignancy, the most commonly selected phrase was “cyst” by radiologists (46%) and “benign cyst” by referring physicians (33%). When asked whether such a lesion required follow-up imaging, 100% of radiologists responded negatively. When asked whether the radiology report should explicitly indicate whether follow-up is recommended for such a lesion (1–5 scale, with 5 indicating the highest level of agreement), referrers' mean response was 3.9 ± 1.2 (median response of 4; 76% responding 4 or 5).


Expressions used to describe potentially low-risk lesions often were associated with greater levels of concern by referring physicians and patients (the latter to an even greater extent) than by radiologists. This discrepancy in meaning is problematic given the frequency of incidental findings and thus of the usage of such expressions within radiology reports. Only a firm statement that a lesion was benign was associated with a median level of concern of 0% by all three groups. Otherwise, all remaining expressions were associated with varying degrees of concern by referrers and patients. Even phrases that radiologists may use when truly believing an incidental lesion to be overwhelmingly likely to represent a cyst (e.g., “statistically likely to be a cyst,” “too small to characterize,” or “most likely a cyst,” all of which were associated with < 1% perceived likelihood by radiologists) had higher perceived risks by patients, thereby potentially increasing patient anxiety. Thus, patients are prone to come away from the report with a different message than the radiologist intended.
These observations indicate the effect of radiologists' choice of language on the ordering of follow-up examinations. Responding radiologists unanimously indicated that lesions with benign imaging features did not require imaging follow-up. However, a referrer's likelihood of ordering follow-up imaging varied markedly according to the expression used to describe the lesion, ranging from 22% to 88% when the lesion was reported as anything other than a “benign cyst.” Expressions incorporating a comment that tumor was “not excluded” were particularly likely to trigger follow-up imaging in most cases. Most referrers preferred guidance within the radiology report regarding the need for following-up imaging. The obtained data indicate that the expression selected by the radiologist for describing an incidental lesion can play a key role in providing such guidance. Although referrers most often favored the phrase “benign cyst” to describe a small incidental liver lesion with benign imaging features, thereby offering a level of certainty, radiologists most often preferred the word “cyst.” The role of specific language within radiology reports in affecting the intensity of subsequent follow-up testing and intervention represents an important area for continued investigation and understanding.
Past literature has also shown variability in the level of certainty that referring physicians associate with expressions of confidence within radiology reports [5, 6]. For example, Hobby et al. [6] noted that, among 11 surveyed physicians, certain phrases (including “absent,” “certain,” and “definite”) were interpreted more consistently among a list of 18 expressions. In addition, Rosenkrantz et al. [5] observed that, among 49 surveyed referrers, certain phrases (including “compatible with,” “suspicious for,” and “possibly”) were associated with particularly poor reproducibility in interpretation. In comparison with these earlier studies, the present investigation formally compares perceptions among three distinct groups (radiologists, referrers, and patients), providing additional insights that are not appreciated by simply showing inconsistency in interpretations among referrers alone. Moreover, the current study provides a more targeted evaluation of language relevant to potentially low-risk findings for which further testing may not be warranted, as well as of the actual associations between such language and follow-up ordering tendencies.
The findings have implications regarding strategies intended to reduce miscommunications between radiologists and their referrers and patients. Although radiologists have historically sought to modify their reporting language in recognition of differing interpretations of a given descriptive term [810], terms used to characterize the level of uncertainty may themselves be a source of further uncertainty [5, 6], as previously noted. Rather, radiologists may adopt alternative approaches to convey their level of confidence within reports [5, 6]—for example, by providing a percentage rather than a verbal expression [5, 6] or establishing a limited set of verbal expressions to be applied by the radiologists in a given practice to correspond with varying levels of concern [11]. Moreover, radiologists may pursue new practice models, such as integrated reading rooms within clinical areas [12], technology-assisted virtual consultation with referrers [13], and diagnostic radiology patient consultation clinics [14], that provide greater opportunities for direct interactions with referrers and patients. Radiologists' meeting with patients and referrers to jointly review the imaging, discuss the findings and their implications, and answer questions greatly enhances communications. Such direct dialogue provides an opportunity to clarify the significance of incidental findings and ensure understanding. Ultimately, regardless of the potential adoption of new reporting or practice approaches, interpreting radiologists must maintain an awareness and vigilance regarding this issue, being mindful of the effect of their selected language on referrers' and patients' perceptions and committing to the use of clear phrasing while avoiding needlessly vague or ambiguous language that may result in greater anxiety or follow-up than is truly warranted for an individual case. Radiologists' consideration of how patients may interpret their report's language is of growing importance as patients are gaining increasing access to their reports through electronic portals [15].
A recent series of editorials by Hillman [16, 17] and Hoang [18, 19] in the Journal of the American College of Radiology focuses explicitly on the need for clearer language and provides perspectives regarding the effect of specific words used in radiology reports. Although controversial [17], the series seeks to address the use of “jargon, clichés, defensive language, and meaningless phrases” that may hinder communication [16]. One particular editorial in the series addressing the reporting of insignificant findings emphasizes the need not to assume that the reader will understand the finding's lack of significance, but rather to note the finding's clinical importance so that the patient and referrer will not be left wondering regarding the need for follow-up [18]. An additional editorial in the series specifically advises avoiding the phrase “not excluded” [19]. Indeed, in the current study, referrers indicated they would order follow-up imaging in response to this phrase in most instances, even if accompanied by additional language indicating that the described lesion is most likely a cyst.
This study has a number of limitations. First, only radiologists, referring physicians, and patients from a single institution were surveyed; survey results could have differed in other centers. In addition, the response rate was low for referring physicians and could not be computed for patients, given the approach of leaving printed copies and fliers in patient waiting areas; thus, it is not possible to know how many patients in fact viewed such materials and considered participating. Also, the expressions were assessed in the context of a small incidental liver lesion; results may have differed as well for incidental findings in other body regions. Finally, as in prior studies [5, 6], the survey was hypothetical in design and did not ask referrers to evaluate actual clinical reports on their patients or ask patients to evaluate their own reports.
In conclusion, referring physicians and, to an even larger extent, patients had greater perceived concern than did radiologists regarding numerous expressions used in radiology reports to describe an incidental finding. Only the expression “benign cyst” was associated with no concern in all three groups, with this expression being referrers' preferred phrase for describing a small liver lesion with benign imaging features in a patient without increased risk factors. Referrers' likelihood of ordering follow-up imaging varied widely among the expressions. Expressions containing the phrase “not excluded” were associated with the highest levels of concern as well as the highest likelihood of ordering follow-up imaging. Because such ambiguity in radiologists' language for describing incidental low-risk findings is a potential source of increased patient anxiety and follow-up testing, greater radiologist attention to the use of such language and potentially new practice or reporting strategies are warranted.


Berland LL, Silverman SG, Gore RM, et al. Managing incidental findings on abdominal CT: white paper of the ACR incidental findings committee. J Am Coll Radiol 2010; 7:754–773
Berland LL. Incidental extracolonic findings on CT colonography: the impending deluge and its implications. J Am Coll Radiol 2009; 6:14–20
Xiong T, McEvoy K, Morton DG, Halligan S, Lilford RJ. Resources and costs associated with incidental extracolonic findings from CT colonography: a study in a symptomatic population. Br J Radiol 2006; 79:948–961
Casarella WJ. A patient's viewpoint on a current controversy. Radiology 2002; 224:927
Rosenkrantz AB, Kiritsy M, Kim S. How “consistent” is “consistent”? A clinician-based assessment of the reliability of expressions used by radiologists to communicate diagnostic confidence. Clin Radiol 2014; 69:745–749
Hobby JL, Tom BD, Todd C, Bearcroft PW, Dixon AK. Communication of doubt and certainty in radiological reports. Br J Radiol 2000; 73:999–1001
Henshaw D, Okawa G, Ching K, Garrido T, Qian H, Tsai J. Access to radiology reports via an online patient portal: experiences of referring physicians and patients. J Am Coll Radiol 2015; 12:582.e1–586.e1
Patel NH, Lauber PR. The meaning of a nonspecific abdominal gas pattern. Acad Radiol 1995; 2:667–669
Bundens WP, Bergan JJ, Halasz NA, Murray J, Drehobl M. The superficial femoral vein: a potentially lethal misnomer. JAMA 1995; 274:1296–1298
Staubesand J, Steel F, Li Y. The official nomenclature of the superficial veins of the lower limb: a case for revision. Clin Anat 1995; 8:426–428
Panicek DM, Hricak H. How sure are you, doctor? A standardized lexicon to describe the radiologist's level of certainty. AJR 2016; 207:2–3
Rosenkrantz AB, Lepor H, Taneja SS, Recht MP. Adoption of an integrated radiology reading room within a urologic oncology clinic: initial experience in facilitating clinician consultations. J Am Coll Radiol 2014; 11:496–500
Rosenkrantz AB, Sherwin J, Prithiani CP, Ostrow D, Recht MP. Technology-assisted virtual consultation for medical imaging. J Am Coll Radiol 2016; 13:995–1002
Mangano MD, Bennett SE, Gunn AJ, Sahani DV, Choy G. Creating a patient-centered radiology practice through the establishment of a diagnostic radiology consultation clinic. AJR 2015; 205:95–99
Bruno MA, Petscavage-Thomas JM, Mohr MJ, Bell SK, Brown SD. The “open letter”: radiologists' reports in the era of patient web portals. J Am Coll Radiol 2014; 11:863–867
Hillman BJ. Speaking of language. J Am Coll Radiol 2015; 12:544
Hillman BJ. Speaking of language revisited. J Am Coll Radiol 2015; 12:1130
Hoang JK. Insignificant findings: don't leave them questioning the significance. J Am Coll Radiol 2015; 12:1244
Hoang JK. Avoid “cannot exclude”: make a diagnosis. J Am Coll Radiol 2015; 12:1009

Information & Authors


Published In

American Journal of Roentgenology
Pages: 140 - 143
PubMed: 27657356


Submitted: April 24, 2016
Accepted: July 4, 2016
Version of record online: September 22, 2016


  1. incidental findings
  2. radiologists
  3. radiology report
  4. standardized reporting
  5. surveys



Andrew B. Rosenkrantz
Department of Radiology, Center for Biomedical Imaging, NYU School of Medicine, NYU Langone Medical Center, 660 First Ave, 3rd Fl, New York, NY 10016.


Address correspondence to A. B. Rosenkrantz ([email protected]).

Metrics & Citations



Export Citations

To download the citation to this article, select your reference manager software.

Articles citing this article

View Options

View options


View PDF

PDF Download

Download PDF







Copy the content Link

Share on social media