Beyond the Blame Game
It is not accuracy of interpretations, or contrast-media administration, or wrong-site surgeries, or infection control in the MRI suite. No, the specialty’s number-one safety issue is radiation exposure, subject of this month’s cover story. In a sadly ironic twist of fate, the very source of the profession’s power of inquiry has become its Achilles heel. A further irony is that the use of software intended to mitigate dose might have exaggerated it. While recent instances of overexposure have been the source of negative attention, they also provide a good argument for keeping the radiologic technologies squarely in the hands of a specialty trained to administer them. Though uncomfortable—particularly because of some newspapers’ reporting inaccuracies and oversimplifications—the crisis is resulting in the further education of patients, referring physicians, radiologists, and technologists. That, ultimately, is a good thing. The impulse might be to circle the wagons, but a better response would be to surround the problem and take ownership on every level. Part of this process is understanding how these unfortunate events could have occurred. J. Anthony Seibert, PhD, president-elect of the American Association of Physicists in Medicine (AAPM) and professor of diagnostic imaging physics at the University of California–Davis, offers some perspective on the dose-limiting software in use at several of the institutions where the overexposures occurred. When used properly, the software compensates for differences in attenuation when imaging noncircular body parts, he explains. If you have a fixed milliamperage, more radiation is transmitted through the thin parts of the body than through the thick parts, resulting in a noise level that is relatively lower in the thin parts than in the thick parts. Seibert says, “What automatic exposure control attempts to do is change the milliamperage interdependent on the angle of projection through the body, so it will be higher through thicker parts of the body and lower through the thinner parts.” The software on the scanner implicated in some of the overexposure incidents is controlled by a noise-index parameter, Seibert explains. “The noise index is a measured variability of the x-ray beam as it penetrates through the body, and in the reconstructed image, it is the standard deviation, within a region of interest, in a relatively uniform part of the image,” he says. “If you want low noise (meaning high dose), you can set the noise index low, and what happens is the radiation will be high to meet the low-noise requirement. Automatic exposure control can be misused, if you don’t know how to use it.” Another potential cause of overexposure would be changing the slice thickness from 5 mm (the common way a CT perfusion study is acquired) to the thinnest possible slice setting, at 0.625 mm—a factor of eight—without making an adjustment to the noise index. “It just so happens that the variance within the reconstructed image from the scanner is going to be proportional to 1 over the slice thickness, and the variance is the square of the standard deviation,” Seibert notes. “Training is important, and understanding and verifying that people are trained also is very important. It does work both ways. It’s not just the manufacturers’ responsibility; it’s a collective responsibility, and there are holes on both sides of the fence.” Organizations, like people, are tested over time, and an appropriate response can help ensure that the test is not repeated. Medical specialties also have been subjected to tests. About three decades ago, a television newsmagazine ran an indictment of the safety record of the anesthesiology specialty, which was in danger of being crushed by exorbitant malpractice fees. The American Society of Anesthesiologists responded by providing seed money to launch the Anesthesia Patient Safety Foundation, admitting physicians, nurses, insurers, and vendors. This collaborative initiative resulted in research/development efforts that produced a patient-safety database, as well as a number of technical and operational safety advances, vastly improving the safety record of the specialty.¹ What, then, is radiology to do? The AAPM held a Dose Summit in May that focused on the impact of the CT dose index volume, as well as other factors in the safe and effective use of CT. It was well attended by all stakeholders, including radiologists, physicists, technologists, vendors, and regulators. The AAPM will continue to offer education at the November RSNA meeting and at the American Roentgen Ray Society’s meeting in the spring. The ACR® has begun work on a national dose registry and is building on its Image Gently campaign. The Medical Imaging Technology Alliance is working, on the vendor side, to arrive at some common solutions. A significant grassroots response also has arisen. You are likely to be impressed when you read about some of the many dose-limiting projects underway in radiology departments across the country. Clearly, many in radiology know what to do with the radiation-safety issue: Own it. Cheryl Proval