Some published articles are meant to be provocative and stimulate discussion, unfortunately on occasion the topic and suggestions are so illogical and irrational, they warrant immediate dismissal. One such paper was published recently in The Journal of the American College of Radiology entitled "Rethinking Normal: Benefits and Risks of Not Reporting Harmless Incidental Findings". The premise is that with the plethora of incidental findings currently being detected by imaging that radiologists should consider the option to not mention them at all.
Many of the premises of the article are inherently flawed. Even to consider not mentioning incidental benign appearing findings is untenable.
The query is made " if such an incidental finding has no known clinical consequence, does it merit mention in a radiology report?" The determination as to whether any finding has clinical consequence is highly contingent on having a detailed and accurate medical history. Every radiologist can attest that we rarely have a comprehensive, accurate history and even in the best of circumstances with complete access to the Electronic Health Record (EHR) we never examine the patient and virtually never go through , family history, review of systems, known environmental exposures , explore the preferences of patients, etc. All findings are of unclear clinical significance until correlated with all the above data elements including labs.
The definition of incidental finding is extraordinarily difficult since the assessment is made only in retrospect by the referring physician who has the most information to come to that assessment. For radiologists, a priori, to dismiss a finding is perilous.
The article fails to recognize the highly variable sensitivity and specificity of certain imaging modalities. The specific example given is that of a Bosniak category 1 cyst. While CT is superb at characterizing cysts how often are CT studies done both with and without contrast to be able to make the determination that cysts are truly not enhancing, have no perceptible wall and no septations. Frankly CT frequently misses very fine septations which maybe very conspicuous on ultrasound.
A failure to recognize the wide range of normal and unique circumstances of each patient also makes the concept of not mentioning "incidental benign findings" impossible. Does a radiologist report a 2.8 cm size of the aorta in a patient being assessed for liver function test abnormalities? The answer is, "it all depends". Do you have a previous study to confirm lack of progression? Does the patient have known atherosclerotic disease? Is there a family history of rapidly progressive AAA ? Does this size represent a 50% increase in size of the remaining aorta which is not visualized on the study? It would virtually never be mentioned if stable for 10 years or if completely uniform throughout its course. We as radiologists most often, do not have all the necessary facts.
The article fails to mention one of the greatest dangers in radiology which is the inability for the radiologist to assess the dynamic nature of a finding. Would the average qualified radiologist not mention a 3.5 cm Bosniak category 1 cyst if three months earlier no cyst was present at all? Does the rapid appearance of what otherwise appears to be a benign cyst warrant comment, likely yes. The article further suggests that lesions below a certain size could possibly be ignored. Unfortunately, the smaller a lesion, the more difficult to accurately characterize it morphologically. Radiologists and doctors typically rely on the low statistical likelihood that small lesions are significant but not based on the intrinsic character of the lesion.
The article relies on access to an EHR and past imaging but frequently there is a limited or inaccessible EHR or previous exams. Often only a written report may be available. If multiple benign appearing cysts are seen in both kidneys but are not mentioned in the previous report, what is the radiologist to conclude? That these are new and rapidly progressive, missed by the previous rad or just not reported.
The medical legal discussion in the paper is based on a tenuous analysis. The current standard of care is to report all findings present, including those thought to be incidental and benign ( all corners of the film). By not reporting incidental findings, radiologists are breaching the standard of care. Not reporting them suggests to most lay people that the radiologist either missed the findings or was too busy to report them. While rare, benign appearing findings do sometimes turn out to be significant and even benign lesions have the potential to malignantly degenerate. Aren't these rare events, with potential catastrophic damages, the exact genesis of lawsuits.
I cannot think of any incidental benign finding which should not be reported. The list is extremely long, atherosclerotic disease, gallstones, diverticulosis, underlying arthritis or spinal stenosis, incidental hernias, hepatic cysts, pancreatic cysts, lung scarring, prostate enlargement, sinus mucosal thickening, ovarian cysts, breast cysts, adrenal lesions, thyroid nodules etc Without complete information about the patient, and understanding the possibility that all conditions may be progressive every finding warrants reporting.
The article concludes by stating that the impetus was to encourage the discussion based on two factors, the plethora of incidental findings rads now face and to enhance patient care. The amazing technology, allowing more detailed characterization of anatomy, has many consequences which include the discovery of more findings of unclear significance. Rather than ignore them, we should make every reasonable effort to identify and report all of them. Further, the greatest disservice we can provide patients and our referring physicians is to, without their input, decide unilaterally what we believe to be incidental and benign without all the facts. It is reminiscent of doctors invoking therapeutic privilege to protect patients from themselves and too much information which could scare them. Patients and their doctors are best positioned to decide what information is relevant and how to best manage it. I mention all the findings and leave to an informed physician in discussion with the patient how to manage their conditions. To do any less is to undermine patients autonomy to decide for themselves.
I do fully endorse and look forward to the day that algorithms will analyze images faster and more accurately than radiologists and that the role of radiologists will be as clinical consultant using Bayes theorem with the ordering clinician to identify the significance of the findings identified by the interpreting computer software.
Until that time radiologists may be best served by bifurcating reports into likely pertinent imaging findings based on the information they have and likely incidental findings to which they defer to the more comprehensive knowledge of the ordering physician to assess significance.
Do not crash and burn. Report all findings recognizing that the vast majority will turn out to be innocuous. Recommend follow up only when there is evidence based medicine to support the practice. Follow the golden rule. Report exams the way you would want your own exam or that of a loved one to be interpreted. Always be an available resource to referring doctors to discuss the significance of findings but then leave management and conclusions in the hands of those best positioned to decide, the ordering physician and most of all the patient.
Nicolas Argy, MD, JD
Copyright © 2016 Nicolas Argy