REPORT STYLE GUIDELINES
1) Introduction
a) As far as possible, a uniform “style” is desirable for our radiology reports, because it prevents discordances in terminology between reports and portrays the section as a group of like-minded radiologists. This guide is a brief outline of desirable common features for our reporting terminology. This guide addresses “style”, as distinct from “content”. The radiology report is a formal medicolegal document that is often the primary means of communication between radiologist and referring physician, and therefore the terminology should be concise, clear, and pertinent. Accordingly, guidelines are arranged under the headings of brevity, clarity, and pertinence, although there is clearly overlap between these areas. The goal of a good radiology report is to provide sufficient detail for the referring physician to generate a “mental picture” of the abnormality, and to know the suspected diagnosis, or the next appropriate step in evaluation.
2) Brevity
a) Avoid beginning sentences with “There is/there are...” such sentences can nearly always be combined with the next sentence(s). E.g., “There is a spiculated mass in the right upper lobe. The mass measures 5 x 4 cm. This finding is suspicious for bronchogenic carcinoma.” can be replaced by “A 5 x 4 cm spiculated mass in the right upper lobe is suggestive of primary bronchogenic carcinoma”.
b) Avoid impressions that are simply descriptive; “a rambling description of findings without a reasonable conclusion does not add anything positive” [1], and also does not prevent successful litigation [2].
c) Avoid redundant phrases, e.g., “review of the scan at bone windows shows no evidence of metastatic disease”. If we report “no bone metastases seen”, then it is implied that we have examined the appropriate windows, in the same way as “the lungs are clear” implies we have reviewed lung windows.
d) “Identified” and other similar terms are frequently overused. E.g., “The right lung nodule is again identified and is unchanged” can be shortened to “The right lung nodule is unchanged”.
e) Long-winded terms for “normal” or “unremarkable” add needlessly to length and may ultimately convey less meaning. For example, “no focal liver mass or intrahepatic duct dilation seen” instead of “the liver is unremarkable” is not only longer, but strictly does not exclude the possibility of a diffuse liver abnormality such as cirrhosis.
3) Clarity
a) Use plain English instead of radiology jargon or abbreviations. E.g., use “center” instead of “epicenter”, “no....” instead of “no significant...", “suggests x” instead “suggests the presence of x”, “lung cancer” instead of “lung CA’, and “metastases” instead of “mets”.
b) The term “is evidence of...” should only be used for findings which are inferred and not directly observed. E.g., “No evidence of portal venous hypertension”. Conversely, it is inappropriate to say “no evidence of pleural effusion”, since the phrase “no pleural effusion” is preferable.
c) Avoid the adjective “significant”. E.g., “No significant adenopathy” – does this mean there is insignificant adenopathy?
d) Use the active rather than the passive tense. E.g., “The pancreatic head mass obstructs the common bile duct”, rather than “The pancreatic head masses causes obstruction of the common bile duct”.
e) The phrase “cannot be excluded” should be avoided as far as possible; it is a grammatically undesirable double negative, and is used differently by radiologists. E.g., stating “spiculated 4 cm lung mass, bronchogenic carcinomas cannot be excluded” when the actual intended meaning is “spiculated 4 cm lung mass, bronchogenic carcinoma is likely”. When a diagnosis is mentioned, but considered unlikely, other options are to state “x is a remote possibility/consideration” – this expresses the intended meaning without using an unwieldy double negative.
f) Do not use terms that have little or no radiological meaning. E.g., “shotty lymphadenopathy” is a clinical term for lymph nodes, which have a “hard granular texture on palpation,” [3] and has no place in a radiology report.
g) Certain terms are interpreted differently by clinicians than by radiologists, e.g., “collection” is often thought of as meaning abscess. Unless this is the intended impression, less loaded terms such as “fluid pocket/fluid-filled structure” may be preferable. Another alternative is to be more explicit, e.g., “Fluid collection, which may be sterile or infected”.
h) Statements such as “careful examination of the retroperitoneum…” may be well intentioned but are harmful, since they could be taken to imply that other body parts are not carefully examined.
i) Certain terms have no uniform meaning between radiologists, so there is little chance that our referring physicians understand what we mean. E.g., “nonspecific bowel gas pattern” is understood by different radiologists to mean either normal, equivocal, or abnormal representing ileus or obstruction [4]. Such terms should not be used alone, but rather the intended meaning should be stated explicitly.
j) Sufficient detail should be provided to allow the person reading the report to visualize the findings. For example, size is a critical feature, which is often omitted. A 1.2 cm retroperitoneal node is different to a 5 cm node. For lesions which are not discrete, a semiquantitative assessment of tumor volume should be provided, although this is necessarily somewhat subjective. E.g., “small volume infiltrative retroperitoneal adenopathy”.
a) Findings of no clinical relevance are of doubtful interest in the body of the report, and certainly do not belong in the impression E.g., “multiple stable renal lesions that are too small to characterize” in a patient with progressive lung metastases. It is unusual for a clinically relevant impression to consist of more than three points.
b) Dependent atelectasis is very common, and does not merit specific mention, unless unusually prominent or thought to represent infiltrate or interstitial disease.
c) The impression should be ordered by the importance of the findings, rather than following an anatomical list. It is disconcerting to read an impression that begins with “small hiatus hernia, gallstones”, and so on, only to end with “new bone metastases”.
d) While extraneous findings are a distraction, it is important to include pertinent negatives. For example, when hydronephrosis is seen, it is important to comment on the presence or absence of renal cortical loss, the level of transition from dilated to non-dilated ureter, and the presence or absence of a stone or visible obstructing mass.
e) When findings are truly nonspecific or equivocal, it is better to offer advice on the appropriate next test than giving a long list of differential diagnoses. From a practical point of view, a differential that includes four or more diagnoses is unlikely to be helpful.
f) The terms “if clinically indicated” or “clinical correlation suggested” should be used sparingly, and should never be used as a substitute to offering a diagnostic opinion. For example, to suggest clinical correlation for a solitary 4 cm solid non-fat containing renal mass is inappropriate, such a mass has to be considered a renal cell carcinoma until proven otherwise, irrespective of symptoms or hematuria.
g) Only additional investigations that are truly likely to be diagnostic should be suggested. For example, suggesting pelvic ultrasound to further evaluate a lobulated heterogenous uterus seen on CT is of doubtful utility, since most likely ultrasound will show a lobulated heterogeneous uterus but will not change the diagnosis of probable leiomyomas.
h) Comparison studies are critical in radiology, and especially in cancer imaging. It is important to state the date and type of prior examination at the beginning of the study, e.g., “Compared to the prior CT of 6/5/00...” It is also good practice to begin the impression with the date of the prior study, e.g., “Since 6/5/00, interval progression of hepatic metastases”.
i) It is not always sufficient to compare the current study to the most recent study, since comparison to earlier studies may demonstrate slow progression of malignant disease, or long-term stability of benign lesions. Not only is this standard of practice we would like for our friends and family, but it may also be medicolegally required [5].
j) The comparison phrase “...was not seen on the prior study” is ambiguous. For clinical and other reasons, it is important to explicitly state if a lesion was or was not visible on the prior study, irrespective of whether or not it was reported.
k) The term “too small to characterize” should be used cautiously. Sub-centimeter hypodense lesions may be too small to characterize by CT density measurements, but comparison to prior studies may allow basic characterization as likely benign or malignant. Sub-centimeter lesions which have been stable for many years are likely benign, while multiple new sub-centimeter lesions are concerning for metastases.
l) False reassurances should be avoided. E.g., “spinal metastases, without evidence of epidural extension” on a CT report disregards the fact that routine CT is not accurate in this assessment, and may give a false impression to the clinician.
References
1. Spira R. Clinician, reveal thyself. Appl Radiol 1996; Nov: 5-13.
2. Berlin L. Malpractice issues in Radiology: Radiology reports. AJR 1997; 169: 943-946.
3. Churchill’s illustrated medical dictionary. Churchill Livingstone Inc, New York. 1989; 1713.
4. Patel NH, Lauber PR. The meaning of a nonspecific bowel gas pattern. Acad Radiol 1995; 2: 667-669.
5. Berlin L. Must new radiographs be compared with all previous radiographs, or only with the most recently obtained radiographs? AJR 2000; 174: 611-615.