It is hard to overemphasize the importance of the radiographic report. The essence of radiologic diagnosis is the correct observation and interpretation of radiographic findings, but even correct observations and interpretations are inadequate if they are not conveyed appropriately. Among the critical elements are information, style, and communication.
It would be wrong to suggest that there is only one correct way to convey what is gleaned from the radiograph. To reinforce that point, two of us have teamed up to air our views on the subject. We begin with comments from Barry Gross.
Let me begin by expressing my understanding of how difficult it is to dictate radiographic reports. My experience indicates that success in the preclinical years of medical school helps to prepare you for success in the clinical years. Success in the clinical years improves your likelihood of success as a clinical intern. However, nothing in my previous experience prepared me for the first few months of my radiology residency. The responsibility for interpretation of images was daunting enough. Even more unforgettable was the literally painful exercise of trying to dictate a small stack of chest radiographs for the first time. I remember sitting with dictaphone in hand, struggling to put into words what needed to be said.
There are two times when it is particularly important to consider the information content of your radiographic reports: when you first start dictating and all the time after that. It is helpful to remember that the referring clinician reading your report may not have images to view at the time the report is read. You should try to put yourself into that clinician’s place, making sure the descriptions you provide will be meaningful even to someone not looking at the radiograph.
In addition, be sure your report says what you intend. We conduct a very instructive teaching conference with our residents from time to time. One resident is positioned in a far corner of the room, unable to view the radiographs being presented. A second resident has to describe the findings on the radiographs so that the first resident is able to make a correct diagnosis. It is eye-opening to the residents to see what erroneous diagnoses their initial descriptors evoke and just as instructive to find out how easily many diagnoses can be made with a few well-chosen phrases. Dictators of radiographic reports should be sure they are choosing phrases in this fashion.
Here are a few questions to consider. Have you made it clear what is abnormal (if there is an abnormality)? Does your description indicate where the abnormality resides? If there is a possible abnormality, have you indicated your degree of certainty that the finding is abnormal? If there is uncertainty, have you suggested how it can be resolved?
I do not mean to imply that a definitive diagnosis can always be made. I also suspect that the more difficult cases for diagnosis are also often the more difficult cases for dictation. A good radiographic report is not always a report that gives the answer; sometimes it is simply a report that clarifies what further questions need to be addressed. However, clarity is an important feature of a good report. When a clinician finishes reading a good radiographic report, there ought to be a sense of understanding about whether the radiograph is normal, abnormal, or possibly abnormal; about what the abnormality does or might represent; and about how remaining uncertainties can be addressed (1).
To coin a phrase, “Brevity is the soul of wit.” Brief accurate reports are far more helpful to clinicians than long accurate reports. Brief reports are far more likely to be read in their entirety. Residents sometimes like to run through long lists of pertinent negatives in their reports as proof that they have looked at a variety of structures in the chest. My advice is to keep the search pattern but do not feel compelled to mention everything you have looked at. Truly pertinent negatives are great (if the clinical history is rule out pneumothorax, it is advisable to say “no pneumothorax”). Otherwise, my personal favorite report is “normal.”
There are a number of ways to organize reports. Some radiologists like to dictate findings and then separately to dictate conclusions. This is certainly an acceptable way to organize a report, but there is a huge pitfall to avoid. A frequent outcome of this reporting style is as follows:
Findings: The heart is enlarged. There are thickened fissures and septal lines as well as vascular indistinctness. There are bilateral pleural effusions. Impression: Cardiomegaly with fissural thickening, septal lines, and pleural effusions.
The problem is that the “impression” is seldom an impression; it is more often a rehash of what has already been said. This style works well if the report above is redone as follows:
Findings: The heart is enlarged. There are thickened fissures and septal lines as well as vascular indistinctness. There are bilateral pleural effusions. Impression: Congestive heart failure.
Dictation pointer: Be brief.
My current preference is to avoid redundancy by dictating only impressions. To encourage brevity I do not adhere to strict grammatical correctness, instead using phrases. The report described above might look like this:
Impression: Cardiomegaly with interstitial pulmonary edema manifested by fissural thickening, septal lines, vascular indistinctness, and bilateral pleural effusions. This most likely indicates congestive heart failure.
Some of my colleagues prefer to dictate findings and impressions for computed tomography (CT) reports because there are usually more findings. I adhere to the impressions-only style for CT, although I add a technique section that describes what area was scanned, what sort of contrast was administered, and what field of view was used (so it can be used again for follow-up studies). A typical example of such a technique section would be as follows:
Scanning was performed from the lung apices through the caudal aspects of the adrenal glands with uncomplicated administration of oral and bolus nonionic intravenous contrast. DFOV (displayed field of view) = 36.
My favorite CT report is the same as my favorite chest x-ray report (as noted above, it is “normal”).
Whatever I am dictating, I try to maintain a balance between appropriate modesty and inappropriate uncertainty. In other words, I am quick to acknowledge that I do not know what every finding represents or what its significance may be in a particular patient (in one memorable dictation, at least to me, I included the phrase, “I have no idea what this is”). Nevertheless, I hate to see uncertainty creep into every report. Referring clinicians already complain (and joke) about the tendency of radiologists to hedge (i.e., to couch their interpretations in so many layers of doubt that they can never be pinned down to a specific diagnosis). Why add unnecessary fuel to this fire?
In particular, I seldom refer to pacemaker lead tips, for example, as being “projected over the right ventricle,” even when I only have a frontal radiograph to evaluate. I believe that for most vascular catheters, we know where the tips are because the catheters are following the course of the vascular tree, just as they have in the other 22,000 catheter placements we previously reviewed. In almost every instance, that course could only be duplicated by a catheter taped to the skin surface, a highly unlikely alternative. I only allow uncertainty to enter my report if the catheter follows an atypical course or if there has been interval development of a potential complication that could indicate catheter malposition (such as mediastinal widening). The same applies to enteric and endotracheal tube placements. I make no similar claims for chest tubes; because the tube is not in a conducting structure like the vascular tree or the gastrointestinal tract, I do not know if it is in the pleura, lung, chest wall, or somewhere else.
With regard to style, I encourage you to find your own voice. I do not believe my personal style is the only acceptable one; it is simply the best one. Still, I do not require all reports dictated by residents under my supervision to be in my personal style. I try to encourage diverse styles as long as they do not include words, phrases, or errors that are anathema to me.
Dictation pointers: Be clear, confident, but not overweening. Correct your reports as needed. Call results frequently.
We must keep in mind the ultimate purpose of having radiologists to interpret various imaging studies. It is not merely to have highly trained and capable interpreters of those images. We move closer to our true purpose when we better understand the nature of the examinations being ordered and their impact on patient care. The key question for any study is “how will this advance the care of this particular patient?” Although I do not advocate sorting through the chest radiographic requisitions for a particular day to decide which ones should and should not be done, I believe that mindset is important and perfectly appropriate for studies such as chest CT and magnetic resonance imaging (MR). My objective is not to refuse to do examinations so that my life will be easier; it is to recognize that our tests are not always completely benign, and even tests that do not endanger the patient may still have significant costs to the patient, especially financial costs. The bottom line is that there must be communication to the radiologist about the motive behind a proposed study, and communication back to the referring clinician is sometimes essential (and surprisingly welcomed by our clinical colleagues) when there is a better way to approach the given problem.
We get closer still to our ultimate purpose when we understand that communication does not end there. The report itself is a critical element in communicating with the clinicians, especially if it is structured properly. Even if your report is brief, accurate, and to the point, consider these questions:
· What if it is lost in the mail (or cyberspace)?
· What if the referring clinician left for a month-long vacation yesterday and will not see your report for quite some time?
· What if everything in your system works well (in which case I hope you will call me to tell me how it is done) and the referring clinician gets your report in 1 to 3 days: How will that delay impact the particular patient whose abnormalities you have detected?
Although the finely crafted radiologic report often suffices to convey what we want it to, that still may not be enough. I try to call reports to referring clinicians at least 5 to 10 times a day. It is obvious enough that this needs to be done with emergent life-threatening findings (such as tension pneumothorax) and equally obvious that it should be our practice for unexpected potentially ominous findings (preoperative chest for hernia repair showing a noncalcified lung mass). There are many times during the course of a radiologist’s day when it is unclear if a report needs to be called or not. My advice is that if you are not sure, call. I call most times when I detect pneumonia in an outpatient. I call most times when a patient with an extrathoracic neoplasm has what I believe is the first evidence of intrathoracic metastases. I call most times when a catheter or tube placement goes astray or may have caused a complication. I call most times when an immunocompromised host has potential evidence of an opportunistic infection.
Ultimately, we need to realize that although many of us no longer do physical examinations or prescribe routine patient medications or discuss findings and prognosis with patients, our real job is to take care of patients. Communication is a critical part of that job. Some of our communication should be with the patient, especially with the patient who is ill and frightened and in pain. We can reduce patient distress by explaining the more invasive procedures that we do (such as percutaneous biopsies) in sympathetic tones and by making sure that we medicate patients appropriately to limit suffering.
Most of our communication will probably be with the referring physicians. We still make a crucial contribution to patient care when we convey important information promptly and directly and when we make sure that the diagnostic workup does not make unnecessary detours caused by lack of knowledge on the part of the referring physicians. That is where we can be champions of patient care.
Dictation pointers: Make recommendations for additional views (oblique, lordotic), additional diagnostic studies, and, above all, comparison with outside old radiographs. Leave the clinician an out (“if clinically indicated”).
An important element of communication is the recommendation of subsequent workup for questionable findings. I believe it is sometimes our responsibility to tell the clinician what should come next. When I find a small noncalcified lung nodule in a patient without old radiographs at my institution, I typically note that comparison with outside old radiographs could be helpful, but if there are none available, chest CT might be appropriate. If there is a questionable nodule adjacent to the right first costochondral junction, I often suggest apical lordotic radiography. In most such instances, I try hard not to tie the clinician’s hands. I specifically try to include the phrase “if clinically indicated” in most such recommendations. This acknowledges that there may be circumstances of which I am unaware (the availability to the clinician of outside medical records, or a patient request for no further tests, as two examples) that eliminate the need for further evaluation. I do not want the subsequent record to imply that the clinician failed to do something when only I thought it was necessary. In summary, my recommendations are to be confident in your approach and direct in sharing your expertise but without treating the referring clinician as a complete idiot.
Although I have hundreds of pet peeves, with regard to the radiographic report I want to emphasize two. The first is the appearance of the finished product. I have on several occasions addressed the following question to our radiology residents: “What is the tangible product of your work?” If you are a carpenter and you build a set of bookshelves, you have an obvious tangible product. If you are an author and you write a collection of essays, you have a similarly obvious tangible product. The tangible product of a radiologist’s work is less evident.
The essence of our work is patient care, but our specific contributions are often difficult to tease out. In fact, it is both amusing and distressing to read patient charts for evidence of our work, even when we have interpreted studies correctly and (unbelievably enough) even when we have communicated our findings directly to the patient care team. There are still occasions where the daily notes, discharge summaries, and subsequent care plans do not seem to acknowledge the findings we have worked so hard to convey to the referring service.
Even the actual interpretation of images does not really result in a tangible product. The most tangible result of what we do is the finished radiographic report. That being the case, why would you want this end product to be sloppy, rife with misspellings, and full of easily correctable errors? I am always on the lookout for superfluous words and incorrect usage. As an example of the former, I virtually never use the word “sided” (right-sided chest tube is not preferable to right chest tube, in my opinion). With regard to the latter, a typical grammatical example I avoid is “the remainder of the lungs are normal.” The subject of this sentence (remainder) is singular, so it really should be “the remainder of the lungs is normal.” This is so awkward that no one can possibly use it; instead, say something like “the lungs are otherwise normal.” A typical anatomic example I avoid is “there is a pneumothorax in the left lower lobe.” As pneumothorax is a pleural abnormality, it obviously occurs outside of the lung. These examples are not intended to be all-inclusive; I suspect that each dictating radiologist can, with a little thought, come up with individual opportunities for improvement.
I will not tolerate even minor stylistic flaws such as split infinitives because I believe they undermine my authority (I believe the clinician reader wonders whether my sloppiness in reporting extends to sloppiness in interpreting). However, I should note that I go through my checkbook looking for the mathematical error if my bank statement is off by 1 cent (not for the 1 cent, just to satisfy my need to know where I went wrong). You may be able to live with a report that says “study of 1-7-02 compared with prior study of December 22, 2001.” Even so, how can you sign off on a report that notes a right lung nodule in the findings but calls it a left lung nodule in the impression? How can your report include an obvious typing or voice-recognition error resulting in a word that is bizarrely out of context? How can your report include comparison with a prior study whose date is still in the future?
It is clear that many reports are signed in haste, with egregious errors included. I used to dictate “short of breath,” but somehow our typists often translated this as “short of breast.” This is the type of error that simply cannot be included in a final report—and yet this is the type of error that does make its way into our “finished products.” (I now invariably dictate “dyspnea” instead.) Would you want a carpenter in your employ to leave similar flaws in the finished product? Our reports have an extended life of their own and will ultimately be the longstanding evidence of our careers in radiology. Please treat them with the care they deserve.
The other pet peeve I want to address is the inability to communicate findings. I am incredibly sympathetic to the problem of reaching the person who can manage the important information you are trying to convey. A brief list of some of the involved problems would include the following: no referring physician listed on the requisition, wrong referring physician listed on the requisition, illegible referring physician listed on the requisition, referring physician out of town, referring physician unavailable by telephone, services changed since the study was ordered, referring physician was only cross-covering, patient was subsequently transferred or discharged, and “someone ordered that study using my name, and I do not know the patient.” I frequently get frustrated, and I do not like to make more than two or three phone calls to communicate a given finding.
However, conversely there are more ways than ever to communicate. You can E-mail, alphanumeric page, leave a voice mail message, or call the referring physician’s cell phone. You can leave a message with the referring physician’s secretary, ensuring that someone will be able to pass the message along. These are all appropriate for messages that should not be lost but need not be communicated right away (such as a noncalcified lung mass in a preoperative patient). For more urgent messages, you can call the nurse taking care of the patient. You can call the resident covering that patient. For outpatients you can call the clinic coordinator from the referring service. You can call the emergency room physician in charge for an emergency room patient. Obviously, the solution you choose will be tailored to the setting in which you work. Just as obviously, there is usually a solution.
Both of these pet peeves boil down to time. I am personally reading more studies per day than ever before. For those of us in academic medical centers, the amount of protected time for research, teaching, and administration is shrinking. Financial pressures in both academic medical centers and private practices are resulting in diminished support—fewer clerks and secretaries result in more nonphysician jobs being done by radiologists. Referring physicians want us there more hours per day than ever before. We still want to give time to our outside interests, and many of us are particularly eager to have time to spend with our families.
So at the end of the day (or throughout the day) it must be tempting to just sign off reports without reading them carefully. It must be tempting to call reports only when we believe we will be sued if we fail to do so. It must be tempting to give up after one attempt to reach a referring physician. (Somehow, it apparently is not tempting to make our reports shorter, which would actually be a good thing!)
I advise you, I urge you, I implore you, do not give in to the temptation. Despite the time pressures that we face, nonradiologists are not sympathetic to our plight—and they are right! We are incredibly well compensated for what we do. Compared with most physicians we work surprisingly short hours and have many vacation and meeting days per year. Our call burdens are generally relatively minor. But that is not really the issue. Even if all our time concerns were valid, that would still not justify shoddy or sloppy work. We are professionals, and we must perform as such. That requires excellent radiographic reports in every sense of the word, with timely communication of important radiographic findings.
Now, for another slant on the radiographic report, I am turning it over to our alternate sponsor, Phil Cascade.
I begin my comments in a somewhat serious vein, by discussing the medical–legal implications of the American College of Radiology (ACR) Standard for Communication: Diagnostic Radiology. After that, I express some of my own “pet peeves” about radiology reporting.
Standards of Practice for Radiology Reporting
The ACR publishes standards of practice that serve as national benchmarks for the performance of radiology-related professional tasks. Trial lawyers often refer to these standards in malpractice cases, both for and against radiologists. The ACR standards delineate the minimum requirements for diagnostic examinations such as chest CT and chest radiography. These standards also outline some of the professional obligations of radiologists, including (by way of example) the standard that describes the responsibilities for covering imaging services in emergency departments. The ACR Standard for Communication sets requirements for the elements of the radiology report and for the communication of the results. I now describe some of the points contained in the communication standard, emphasizing issues I believe have the most impact on the way we practice. Examples of case law will be cited as illustrations. These cases are abstracted from an article published by Cascade and Berlin (2), with permission from the American Journal of Roentgenology. I encourage the readers to become familiar with the ACR communication standard in its entirety. The publication can be obtained directly from the ACR (3).
The report should, when appropriate, identify factors that may limit the sensitivity and specificity of the examination. Standard Item II.C. 3
A patient fell from a truck and ended up in an emergency department complaining of neck pain. A cervical spine series was done that did not clearly show the junction of the cervical and thoracic spine. The examination was reported as normal, but the limitation of the study was not mentioned. Follow-up spine radiographs were obtained later because the neck pain persisted. A diagnosis of a fracture dislocation was made at C7-T1, and a lawsuit was filed.
The initial jury verdict found in favor of the radiologist. However, the verdict was reversed on appeal, specifying that the radiologist had the obligation to mention limitations of a study when there could be an impact on patient care.
Comparison with previous examinations and reports should be part of the radiologic consultation and report when appropriate and available.Standard Item II.C.5
A radiologist reported the presence of “questionable indeterminate calcifications” in the left breast of a 60-year-old woman. A prior mammogram was not available, although the typed report was. The prior report described the presence of benign-appearing calcifications, and for this reason the radiologist stated that the calcifications on the new study were probably benign. A subsequent screening mammogram 1 year later revealed changes in the calcifications that on biopsy showed carcinoma in situ.
At trial, an expert witness criticized the radiologist for not attempting to obtain the earlier mammogram for a comparison and to evaluate for interval change in the calcifications. The jury found in favor of the plaintiff, awarding her more than $100,000.
The timeliness of reporting any radiologic examinations varies with the nature and urgency of the clinical problem. The final report should be made available in a clinically appropriate timely manner. Standard Item III.A.
A venous hyperalimentation catheter was introduced into a young woman who suffered from chronic Crohn disease. A radiologist reported a concern that the catheter was positioned more medially in the mediastinum than usually seen. The radiologist tried unsuccessfully by phone to reach the surgeon who had placed the catheter. The radiologist did reach a nurse caring for the patient. She assured the radiologist that the catheter had free backflow and that the patient was asymptomatic after hyperalimentation. Hours later, the patient went into shock and died. Autopsy revealed that the catheter had perforated the superior vena cava with extravasation of hyperalimentation fluid into the thorax.
The family of the deceased sued, claiming that it was the responsibility of the radiologist to communicate directly with the surgeon in a timely fashion so that steps could be taken to remedy the problem. A radiologist expert witness testified that the defendant had breached the standard of practice by abandoning attempts to contact the surgeon directly and immediately when perforation was suspected. The case was settled before trial.
The final report should be proofread to minimize typographical errors, deleted words, and confusing or conflicting statements. Standard Item III.B.
A 32-year-old woman had chest radiography to evaluate for pneumonia. A radiologist reported the study as being within normal limits. Follow-up chest radiographs 18 months later revealed a right apical mass that turned out to be a neurofibroma. In retrospect, the mass had been present but missed on the initial chest radiographs.
A lawsuit was filed for missing the lesion on the initial chest examination. The radiologist named in the suit claimed that the report was not his and that the transcriptionist was in error when she put his name on the report as the dictating radiologist. The defendant did testify that it was his signature that was on the report but he had signed the report by mistake. It was his routine to sign all of his reports before going home in the evening, but he did not read the reports very carefully. The associate radiologist denied that he had dictated the report. The case was considered indefensible because both radiologists had accused each other of interpreting the study. The case was settled for $37,600.
In summary, the standards of practice for radiologists as developed by the ACR can be, and are, used by trial lawyers. All radiologists should be familiar with the standards that apply to radiology reporting. Sound risk management mandates close adherence to these standards for medical–legal reasons and, even more importantly, for good patient care.
More Pet Peeves
As Mark Twain once wrote, “It were not best that we should all think alike; it is difference of opinion that makes horse races” (from Tragedy of Pudd’nhead Wilson). In my experience, I found that reporting is one of the most controversial and passionate subjects that radiologists argue about. It seems as though each of us has his or her own style of dictating cases, a style thought (by each individual) to be the best. In the following section I point out examples of errors of composition and use of language and make recommendations for improvement. You can reject the thoughts, or adopt them if you are wise.
Do you remember your English composition classes in elementary school? From what I have seen, most radiologists have forgotten everything they learned. Basic principles of composition such as 1) use the active voice, 2) omit needless words, and 3) put statements in a positive form come to mind.
A typical chest radiographic report might read as follows:
Findings: The lungs are well inflated. There is a calcified nodule in the right lower lobe that is compatible with a granuloma. The lungs are somewhat overexposed and nodules cannot be excluded. The heart is normal in size. There is no evidence of heart failure. There is no pleural effusion. The ribs and spine are normal for a patient of this age. There is a 5 mm osteoma in the proximal humeral metaphysis on the right. Impression: There are no significant findings on this somewhat limited examination (see paragraph above).
How boring and difficult to read! For some reason, radiologists continue to dictate in the passive voice, put in unnecessary words, and on occasion put sentences in a negative form. Hedging is common. I say, use the active voice, do not be verbose, be confident in your observations, and do not use the negative form. To ask the clinician “see paragraph above” or “correlate clinically” is unnecessary and insulting. Reading the body of the report is usually superfluous, and of course the clinician will correlate the report with his or her clinical knowledge.
Another way to write the report could be as follows:
Findings: Fully inflated lungs. Benign calcified nodule in the right lower lobe. Overexposed images. Neoplasms could be present and obscured. Normal heart size, and no failure. Conclusion: Negative except for right lower lobe granuloma.
This report shows brevity, confidence, and clarity. I removed unnecessary words such as the description of the bones and insignificant findings such as a tiny osteoma, and I use phrases rather than full sentences to make the report more succinct. I took out hedges such as “there is no evidence of” I changed the double negative “cannot be excluded,” replacing it with the positive statement “could be present.” I also took the liberty of ending the report with “Conclusion” rather than “Impression.” Conclusion portrays confidence.
An even better report, in my opinion follows:
Conclusion: Negative except for right lower lobe granuloma.
This is my favorite type of report. There is no need to dictate findings; just a conclusion will suffice unless the findings are so complex that a full description is warranted. My expression is “few words, clear message.”
In addition to poor grammar, radiologists often misuse terminology when reporting. The Fleischner Society, dedicated to excellence in chest radiology, has made comprehensive recommendations on appropriate and inappropriate terminology in chest radiology (4,5). Following is a list of terms related to chest radiography that should be banished, in my opinion, from radiology interpretive reports:
· Diaphragms: Humans have one muscular diaphragm, not more. The singular form diaphragm should be used.
· Adenopathy: Adeno comes from the Greek word meaning gland. There are no glands in lymph nodes! Use lymph node enlargementinstead.
· Density: Many reports contain the term density as a descriptor of an abnormal finding in the lung. In fact, density refers to the degree of darkening of the chest radiograph. Abnormalities in the lung such as nodules are relatively white on radiographs, not dark. Useopacity instead.
· Bony: Radiologists often use the adjective bony when describing features related to the thoracic skeleton. For example, I have read reports of bony destruction of ribs. Would these same radiologists speak about footy pain instead of foot pain if they drop something on their foot or lungy congestion instead of lung congestion if they have an upper respiratory infection? Of course not. Therefore, use boneor osseous when describing abnormalities of bone (bone destruction not bony destruction).
· Prominent: A frequent and classic hedge word used by many radiologists. For example, prominent aortic arch comes to mind. What is meant is that the aorta is, or might be, somewhat dilated or enlarged. In general, if you believe a structure is abnormal, say so. If you are not sure, do something to find out. In this example, call the aorta enlarged or recommend further imaging.
· Nonspecific: The term nonspecific often appears in reports describing a bowel gas pattern. I have also seen the same term used to describe the interstitium of the lungs. When used, the term usually means that the radiologist believes the study is probably normal but is hedging his or her bets. Earn your paycheck. Do not dodge the decision by leaving the findings as nonspecific. Call the lungs normal, abnormal, or possibly abnormal with a recommendation of further imaging to find out.
· Mainstem: Mainstem is an archaic term that should be replaced by main as in right main bronchus.
· Poor Inspiration: Do not blame the patient. It could be that the technologist took the image during expiration.
· Low Lung Volumes: Use lung volumes to refer to abnormal lungs. For example, patients with emphysema have large lung volumes and patients with pulmonary fibrosis can have small lung volumes. It is preferable to use degrees of inflation to describe the degree of aeration of the lungs, for example, poor inflation or hyperinflation.
1. Chapman WW, Fiszman M, Frederick PR, et al. Quantifying the characteristics of unambiguous chest radiography reports in the context of pneumonia. Acad Radiol 2001;8:57–66.
2. Cascade PN, Berlin L. American College of Radiology standard for communication. AJR Am J Roentgenol 1999;173:1439–1442.
3. American College of Radiology. ACR standard for communication: diagnostic radiology. In: Standards. Reston, VA: American College of Radiology, 2000:1–3.
4. Tuddingham WJ. Glossary of terms for thoracic radiology: recommendations of the nomenclature committee of the Fleischner Society.AJR Am J Roentgenol 1984;143:509–517.
5. Austin JHM, Muller NL, Friedman PJ, et al. Glossary of terms for CT of the lungs: recommendations of the nomenclature committee of the Fleischner Society. Radiology 1996;200:327–331.