CT utilization has exploded in recent decades, and with it has come an increase in the amount of radiation the population is exposed to annually. While concerns about radiation dose are becoming more publicized, much work remains to fully understand the risks of radiation in medical imaging, according to the presenters in a webinar on Jan. 25 hosted by the Society for Imaging Informatics in Medicine (SIIM).
The benefits of CT, which include non-invasive interrogation, excellent diagnostic capabilities and more powerful scanners, ushered in a revolution in medical care, said J. Anthony Seibert, PhD, of the Imaging Research Center at the University of California–Davis Medical Center. As the modality has expanded, though, concerns about dose have become more prominent in the media and everyone now wants to measure dose information.
Current dose reporting methods use the CT dose index (CTDI) to provide dose comparisons for scan protocols. The CTDI is useful in benchmarking, but not great at estimating the dose received by an individual patient, according to Seibert. Other measures of dose include the volume CTDI (CTDIvol), a conversion that attempts to correct for differences in patient size and shape; dose length product (DLP), which multiplies the CTDIvol by the scan length to indicate the dose imparted to a patient; and effective dose, which is a whole body dose estimate derived from the DLP.
This issue is that CTDIvol and DLP are not actual patient doses, but values generated from the CT scanner using calibration phantoms of 32 cm or 16 cm in diameter, according to Seibert. If the diameter of a patient’s body is different from the calibration phantom, this can lead to significant overestimations or underestimations of the patient dose.
“You cannot blindly use dose indicators without some knowledge of what they mean,” said Seibert.
An unclear understanding of radiation dose is problematic because of the risks associated with ionizing radiation. The webinar’s other presenter, Rebecca Smith-Bindman, MD, professor in the departments of radiology & biomedical imagining; epidemiology & biostatistics; and obstetrics, gynecology & reproductive medicine at the University of California, San Francisco, explained that an estimated 2 to 4 percent of new cancers will be the result of CT radiation exposure.
Smith-Bindman illustrated the increase in radiation dose on a per capita basis. In 1985, the U.S. population was exposed to an average of 3.7 mSv/year, with only a quarter of that exposure coming from imaging and the rest coming from natural sources such as radon. In 2006, per capita exposure was 6.2 mSv/year, with a 50/50 split between natural and imaging sources. CT represents about two-thirds of the cumulative dose from x-ray imaging.
“I’m not saying using CT this much is good or bad, it’s just gone up a lot,” said Smith-Bindman.
Smith-Bindman also pointed to the variability of administered dose between facilities and between physicians. A retrospective study of four San Francisco Bay Area medical centers conducted by Smith-Bindman and colleagues showed a 13-fold variation between doses for each of the study types measured.
High-dose studies which improve image quality are also increasing in use, and increasing radiation exposure as a result. A separate study of two million health maintenance organization members found that by 2010, 16 percent of patients imaged had an annual dose greater than 20 mSv and nearly 5 percent had an annual dose of greater than 50 mSv. For comparison, survivors of the bomb dropped on Hiroshima received an average dose of 40 mSv.
In addition to the increase of high dose studies, Smith-Bindman said radiation dose is so variable because of a lack of clear standards for radiologists. There is no professional or governmental organization responsible for collecting and reporting dose data, and unlike many other developed nations, there are few dose targets set for CT use in the U.S. Patient weight is only a small part of the variation; weight can lead to a twofold variation in dose while protocols and other factors can lead to 100-fold variations.
To help standardize doses across the University of California system’s campuses and optimize diagnosis, the University of California Medical Centers have begun a collaborative project called UC DOSE (Dose Optimization and Standardization Endeavor). The end goal will be a standard set of CT protocols to be made widely available. In February