From Nature:
People may volunteer for a study simply to advance science, but a large fraction of them could wind up receiving unnerving news. A paper published today1 reports finding that 40% of participants in imaging experiments had clinical anomalies beyond the scope of the investigation, and that, of these cases, 6% provoked subsequent medical intervention.
Radiologists at the Mayo Clinic in Rochester, Minnesota, appraise images from research examinations daily and report any potential problems that they spot to physicians. An expert panel of physicians, radiologists and bioethicists assessed the benefits and burdens of radiologists' findings for research examinations taken over three months in 2004 by studying individuals' medical records over a follow-up period of three years. Out of a total of 1,426 examinations, 567 revealed at least one anomaly, and the total tally of anomalies across this subset was more than 1,000.
More here. (Note: For my radiologist sister Ga who has told me of this phenomenon for years!)