Dr. Berlin is at Skokie Hospital, Skokie, IL, Rush Medical College, and the University of Illinois, Chicago, IL.
How do you solve a problem like Maria Incidentalomas?
When I’m with her, I’m confused,
Out of focus and bemused.
(With apologies to) Oscar Hammerstein, Sound of Music, 1959
We know certain facts about incidentalomas. We know what they are: incidentally discovered masses or lesions, detected by computed tomography (CT), magnetic resonance imaging (MRI), or other imaging examinations performed for an unrelated reason.1 We know why incidentalomas are increasing: The number of CT exams performed in the United States has increased geometrically over the decades, rising from 3 million annually in 1980 to close to 80 million annually currently, along with remarkable improvement in spatial and contrast resolution on newer-generation CT scanning equipment.2 We know that up to 70% of persons undergoing screening CT colonography have at least one detectable incidentaloma.3 We know incidentalomas are found in 34% of hemodynamically stable blunt trauma patients.4 We know that 35% of patients undergoing CT for thoracolumbar blunt trauma injuries harbor incidentalomas.5 We know that nodular incidentalomas are found in at least 25% of patients undergoing chest CT.6 We know that incidentalomas occur in at least 40% of abdominal and pelvic CT exams obtained for research purposes.7 We know incidentalomas are present in 49% of patients undergoing aortoiliac CT angiography prior to aortic valve repair.8 We know that incidentalomas are found in up to 50% of the lungs on CT exams of the chest, up to 15% in the kidneys and liver on abdominal CTs, and up to 67% in the thyroid gland on neck ultrasound exams.9 And we know that the chance that an incidentaloma found in any of these exams could represent a lethal carcinoma is < 1%.9
We also know that radiologists are increasingly facing a dilemma: If there is reasonable belief that the incidentaloma is of no clinical significance, then mentioning it in a radiologic report will lead to a cascade of costly tests, sometimes resulting in serious complications. However, if radiologists decide not to mention the incidentaloma, and, in the unlikely event it later turns out to have been an early carcinoma and the patient’s health has been jeopardized, medical malpractice litigation could well ensue.
But what we do not know is: How do we solve the problem of incidentalomas?
Should we ignore incidentalomas?
Many radiologists and nonradiologic physicians believe we should not report incidentalomas if they appear benign.9 As pointed out by 2 medical researchers, “Tests that provide information about unrelated conditions leave the physician and patient to contend with information they had not sought but which they find impossible to ignore; patients would be better served if physicians limited their access to unsolicited diagnostic information.”10
Similar sentiments were echoed by a gastroenterologist: “It defies a basic tenet of medicine to question the benefit of diagnostic information, as any new piece of information regarding a patient’s health is considered valuable and worthwhile having. But acquiring new knowledge can worsen a patient’s overall well-being by leading to more aggravation, discomfort, and injury than the earlier state of ignorance. Even if the patient and physician decide to ignore the positive test, they may still harbor lingering doubts and fear that a potentially serious disease will go untreated. The physician may be concerned about being blamed for missing an important diagnosis and face future litigation.”11
There is yet an additional problem with reporting every incidentaloma: overdiagnosis, defined as “diagnosis of an illness, disease, or cancer that never becomes symptomatic, is not fatal, and may continue until the individual dies from other causes.”12 In other words, radiologic exams have the potential to do harm by finding tumors that would not otherwise have required treatment, setting off a cascade of unnecessary tests and biopsies, sometimes resulting in complications. “It is like the lottery: In exchange for those few who win the lottery, there are many, many others who have to pay the price in human costs.”13
The patient’s right to know
Ninety-eight years ago, the Honorable Judge Benjamin Cardozo issued a dictum that became the foundation for informed consent between patient and doctor: “Any human being of adult years and sound mind has a right to determine what shall be done with his own body.”14 This judicial pronouncement was incorporated into the American Medical Association’s Code of Ethics: “The physician’s obligation is to present the medical facts accurately to the patient… . Physicians should disclose all relevant medical information to patients.”15
The quandary as to whether to disclose a radiologic finding that has an extremely low likelihood of having an adverse effect on a patient’s health was the subject of an editorial written by radiologist Roy Filly. Referring to the very high percentage of false-positive markers indicative of Down syndrome found all too frequently on prenatal ultrasound examinations, Filly wrote:
“The identification of these ‘abnormalities’ in low-risk women has crossed the line of ‘more harm than good’… For the tiny number of Down syndrome fetuses that potentially may come to light by chasing down every last marker, we put at least 10% of all pregnant women with perfectly normal fetuses through a great deal of worry…. Should I simply ignore these features? I wish I had the courage, but I don’t. Even with my considerable clout in the world of obstetrical sonography, I cannot unilaterally ignore them. This is not how American medicine works.”16
In her novel Handle with Care, author Jody Picoult created a fictional scenario in which an obstetrician who, while performing a prenatal ultrasound, observed a questionable marker for osteogenesis imperfecta, but because of the high false-positive rate of the marker, decided not to inform the patient. Later, the mother delivered a baby with the disease and subsequently filed a medical malpractice lawsuit against the obstetrician, claiming that she would have aborted the fetus had the obstetrician informed her of the abnormal finding. In his closing argument at the conclusion of the fictional malpractice trial, the plaintiff’s attorney told the jury, “This case is about facts that the obstetrician knew, but didn’t give the patient. The obstetrician did not cause the illness, but is to blame for not giving the family all of the information. When a physician withholds information from a patient, that’s malpractice.”17
Emphasizing that patients should be informed of every bit of information that could adversely affect their health, even information that is highly unlikely to be injurious, a New York internist asserted, “My patients want to know if they have cancer as early in the process as possible so that they can be treated. They don’t want to have to rely on mathematical projections or statistics about ‘hypothetical’ death rates.”18
Standard of care
The standard of care to which physicians must adhere is defined as conduct that is “usual and customary in the local or national community, under the same or similar circumstances.”19 In formulating how radiologists should conduct themselves when they observe an incidentaloma, we must attempt to discover, usually by consulting the published scientific literature, whether there is a “usual and customary manner” in which other radiologists and medical facilities deal with incidentalomas. A perusal of medical literature reveals that there is no clear-cut unanimity on this issue. A recent report describing how radiologists at 3 major academic centers—Johns Hopkins University, New York University, and Stanford University—manage incidentalomas disclosed that there is anything but a consensus among them. The rate of agreement as to whether to report an incidentaloma seen on a CT scan ranged from 30% to 85%.20 Not only was there lack of agreement among the academic institutions but also among radiologists in the same institution. A report from the University of Pittsburg disclosed that even when incidental findings that were not considered important enough to require medical follow up were discovered, patient notification was nevertheless almost always carried out because of “medico-legal concerns.”21
Other articles have focused on the ethical issue of whether physicians who discover an incidentaloma on imaging studies of volunteers involved in research studies should or should not divulge that information to the volunteer. Not surprisingly, there was no consensus.22,23
The results of a recent retrospective study of patients diagnosed with renal cell carcinoma in Iceland between 1971 and 2005 could well have major impact in determining the standard of care relative to the reporting of incidentalomas. In 28% of patients, the carcinoma was discovered as an incidental finding on a CT or ultrasound performed for reasons unrelated to the renal carcinoma. These patients had a statistically significant better survival rate than those whose carcinomas were found after they had exhibited symptoms or laboratory abnormalities. The authors concluded that incidental detection affects survival favorably—indeed, to a much greater extent than can be explained simply by differences in stage, grade, or demographics compared to those with symptoms.24
In an editorial titled, “Incidentally, It’s Still Cancer,” a surgeon-researcher opined:
“The most important take-home point is that malignancy does occur in small tumors and also can be evident ultimately in presumably benign lesions. Physicians should be applauded for endorsing an aggressive surgical stance against the disease. After all, it’s cancer we are dealing with here.”25
There has as yet not been much malpractice litigation focusing on a radiologist’s failure to report an incidentaloma. Thus, the standard of care that will be applied in such cases cannot be predicted with any degree of certainty.26 Nevertheless, one such lawsuit that did proceed to trial might provide a hint. In the case, a 47-year-old man presented to a hospital emergency department (ED) with flank pain. In the CT scan report ordered by the ED physician, the radiologist stated that the exam was normal except for a “hypodense, well-circumscribed mass in the left kidney, most likely a cyst.” One year later the patient was diagnosed with renal cell carcinoma; he ultimately died. A lawsuit was filed by the patient’s family, and at trial the following interchange between the plaintiffs’ attorney and the defendant-radiologist took place:
Q: Doctor, why didn’t you mention the potentially abnormal finding in the impression?
A: Because I thought the finding was almost certainly of no significance and would have led to a number of unnecessary and possibly dangerous tests.
Q: Could it have represented an early cancer?
A: Yes, but probably no more than a 1% chance.
Q: Doctor, in this case it was 100%. Shouldn’t you have let the patient and his private physician decide whether further testing was indicated? Did you not deprive the patient who is now dying of cancer, rather than cured and living, of his inalienable right to make his >own decision about his health?
The jury rendered a verdict in favor of the plaintiff. Why the jury determined that the radiologist was liable for failing to call attention to the incidentaloma is not difficult to understand. The era of “medical paternalism” (“I am the doctor; I know what is best for you.”) is long past. Radiologists and nonradiologist physicians now live in a consumer-driven society, where physicians no longer make unilateral and arbitrary decisions regarding a patient’s health; rather, they are a partner and an advisor to the patient.27 Patients expect, and indeed want, to be informed of any potential laboratory or imaging abnormality that could possibly adversely affect their health, even if the probability that the abnormality could be injurious is highly unlikely. An example of this attitude was shown on a CBS television news program on April 1, 2012, that focused on a 19-year-old man who suffered cardiac arrest and died suddenly due to a malfunctioning cardiac monitor/pacemaker. Data revealed that the frequency of malfunction in such devices is 0.1%. During the TV interview, the father of the deceased man angrily lamented, “Why weren’t we told that the device could fail? We should have been.” While an adverse event with an incidence of 0.1% should be disclosed to a patient may sound unnecessary and unreasonable to physicians, it appears quite necessary and reasonable in the minds of the news media and the public.
Ultimately, radiologists must decide for themselves how to solve the problem of incidentalomas. One suggested approach would be to call attention to the finding; eg, by stating, “An incidental finding of a 5-mm nodular density in the liver is noted. The likelihood that this represents significant pathology is extremely remote.” In this manner, the radiologist is expressing the opinion that there is a very low likelihood of serious disease, but leaving the decision on what, if any, follow-up diagnostic or therapeutic procedures to pursue up to the referring physician and the patient.