CT Radiation: How Low Should You Go?
What is the right amount of ionizing radiation needed to create diagnostic-quality medical images?
In a new JACR editorial, Mervyn D. Cohen, MBChB, MD, a retired pediatric radiologist, challenges us to consider that the ALARA concept (meaning as low as reasonably achievable) and, by extension, the Image Gently campaign, have outlived their usefulness. Cohen starts by explaining that low radiation dogma may be less impactful as CT vendors are aggressively competing to use computing power to maintain image quality while significantly reducing radiation dose, which may mitigate concerns about the overall amount of radiation from medical imaging. From there, he goes a step further.
Cohen contends that patients and radiologists have been harmed by this over emphasis on radiation dose. He describes what dose-monitoring programs achieved, both positive and negative. Please read this article in its entirety to form your own opinion, but do read it. It never hurts to challenge our assumptions, no matter if it changes our viewpoint completely, a little, or not at all.
A few weeks ago, I saw firsthand examples of Cohen's assertion that nondiagnostic CT studies are an undesirable outcome of the competitive sport of radiation lowering. A brand-new CT scanner had just been installed at our institution, with promises of lowering patients' radiation dose. Almost immediately, I had to put the brakes on chest CTs at our institution. The image quality was suboptimal, with too much image noise. A minor component of the image quality had to do with filter kernel selection, but the applications specialist acknowledged that some of the image quality degradation was absolutely related to radiation dose. After the necessary adjustments, our radiation dose for chest CT is still lower than on our older scanners. And I breathe easily knowing that I am not missing subtle pathology and not hallucinating pathology where none exists.
As I approached the situation at my institution, I had to ask myself: Was I simply used to images performed with higher-than-average radiation? Was that why I needed to have the protocols tweaked? I asked for the CT supervisor to quickly pull ACR Registry Data. Our institution's median CTDIvol (a phantom-standardized measure of CT scanner radiation output) and DLP (dose length product, i.e., CTDIvol times the imaged z-axis length) were lower than those for all comparison groups. So our facility does quite well with respect to keeping radiation dose down.
But with our new CT scanner, we could go even lower. Working together with applications specialists, we found the right balance: we reduced radiation dose, but with comparable image quality. Ours were a little higher than the very low radiation protocols initially suggested, but we maintain confidence in our ability to discern pathology (such as tiny ground-glass pulmonary nodules) from non-pathologic anatomy and to discern true pathology from pseudo-pathology (like increased image noise simulating tiny ground-glass pulmonary nodules).
"We must place patient care first. Our first obligation to our patients is an— Mervyn D. Cohen, MBChB, MD
Cohen and I agree with the most dogmatic of all medical dogma. As Cohen writes, "We must place patient care first. Our first obligation to our patients is an accurate diagnosis." And sometimes that requires us to use CT equipment that delivers ionizing radiation. I think in this regard, ALARA is a reasonable principle, as long as the emphasis is on the "R" (reasonably). If you skip the "R," then you get ALAA (as low as achievable), a concept driven purely by the fear of the harm of radiation. Keeping the "R" in ALARA acknowledges that ionizing radiation causes DNA damage (with theorized future risk of cancer) but also includes the amazing ability of radiologists to quickly answer a vast array of medical questions with CT. The same radiation that sparks so much fear makes it possible to avoid unnecessary exploratory surgery, a riskier proposition. As long as CT is in fact the imaging modality of choice, the risk for patients is lower than the danger of not getting the CT and not getting a diagnosis or getting the wrong diagnosis.