Medical studies often reveal far more information about a participant than researchers seek. So, extraneous information is often ignored. But what happens when researchers notice an unexpected mass in a brain scan or threatening clots in a computed tomography (CT) image? The ethics are still fuzzy about how much responsibility—and freedom—medical researchers should have in informing subjects or their doctors if something unsought is revealed in the course of an imaging study.

A new analysis found that about 40 percent of research imaging scans completed at the Mayo Clinic in Rochester, Minn., included these so-called incidental findings. Very few (6.2 percent) ended up receiving clinical follow-up, but some imaging techniques, such as abdomen and pelvic CT scans, sent nearly 10 percent of participants on for additional assessment after abnormal results. The scans study was published online September 27 in Archives of Internal Medicine.

"We knew that incidental findings were common, but I was a little surprised that they were present in almost 40 percent of the research participants," says Nicholas Orme of the Mayo Clinic and a co-author of the new study.

Despite the imaging these subjects received by participating in a study, extra follow-up was of unclear benefit for about two thirds of the few participants who ended up receiving it. Even though further testing was infrequent, and unnecessary procedures and other burdens to patients occurred in only 5 percent of people with an incidental finding, "clear medical benefit to an individual participant rarely occurs," Bernard Lo, director of the Program in Medical Ethics at the University of California, San Francisco Medical Center, wrote in an essay in the same issue of Archives.

Of the 1,426 images taken for medical research at the clinic in the first three months of 2004 and assessed for the new analysis, researchers identified a total of 1,055 incidental findings—often finding more than one in a single scan. These discoveries were passed along to a participant's doctor, and "the decision on how much to disclose to the patient depends upon the primary care physician," Orme explains. Many institutions, however, do not have a policy of notifying a participant or their doctor or are unable to analyze the images until after a potential intervention window has passed.

Although "people want to do the right thing" for individuals participating in studies, there are also big funding and workload questions to consider, says Susan Wolf, a professor of law, medicine and public policy at the University of Minnesota Law School and paper co-author.

Many studies are designed to assess a narrow range of metrics and often do not make use of the latest diagnostic equipment and analysis. And images are often assessed by someone with background and training specific to the research goals, not a general diagnostic radiologist, Wolf points out. So to ask researchers to also keep an eye out for miscellaneous findings—and then to obtain a specialist's opinion when necessary—requires extra time, diligence and money.

Additionally, "the roles of research and clinician should not be conflated," Lo noted in his essay. "The primary objective of researchers is to obtain valid, generalizable scientific knowledge," he wrote.

To better handle incidental findings when they arise, "researchers have to anticipate this problem," Wolf says. And with more—and better—imaging technology proliferating in the research field, the frequency of incidental findings is likely to increase, she notes, making it "a significant issue."

Participants also might have a skewed view of what kind of evaluation they will receive. "Research subjects may just assume that if there is something wrong with their brain, the researchers will find it"—and tell them, Wolf says. Other participants, however, might also prefer not to be notified of any extraneous findings in their reports. More explicit discussion with potential participants in the consent process could help researchers and participants establish guidelines before examinations begin.

Imaging studies provide many clear cases of a brain tumor discovered early or a questionable-looking mass that turned out—only after surgery—to have been benign and harmless. But the problem of incidental findings proliferates in other kinds of medical research as well, including genetics and genomic investigations. And more work remains to be done on other ramifications of unintended findings. The new study did not take into account more subjective participant experiences, such as whether finding out about an anomaly revealed in a scan caused undue stress or excessive personal expense, which, Lo concluded, "underestimated the net burdens."