Top 10 CT radiation producers 'exposed' in JACR analysis

Among some of the most commonly ordered CT exams, the one that tends to deliver the largest radiation doses is chest abdomen pelvis with IV contrast. The one with the smallest radiation footprint is chest without IV contrast.

These findings came as no surprise to the researchers who analyzed three years’ worth of data in ACR’s dose index registry, but the numbers may help standardize informed improvements in CT protocols.

At least, that’s the hope of Joanna Escalon, MD, of the radiology department at New York Presbyterian-Weill Cornell Hospital, and Mythreyi Chatfield, PhD, director of data registries at ACR, and colleagues.

Their study is running in the August edition of the Journal of the American College of Radiology.

The authors looked at 5.9 million CT scans entered into the registry from 616 different facilities between July 2011 and July 2014.

The sites spanned facilities performing fewer than 100 scans per month to those doing more than 3,000.

Using as their primary measure dose length product (DLP), a calculation that estimates a scan’s dose to a given area irradiated, the researchers crunched the numbers on 10 familiar CT scans.

Here they are, listed in order of greatest to least radiation (expressed as DLP in milligray units per centimeter): 

1. Chest abdomen pelvis with contrast—mean DLP 1322 mGy-cm (median: 1176)
2. Head/brain without contrast—mean DLP 987 mGy-cm (median: 876)
3. Abdomen pelvis with contrast—mean DLP 938 mGy-cm (median: 773)
4. Abdomen pelvis without contrast—mean DLP 852 (median: 745)
5. Abdomen pelvis kidney without contrast — mean DLP 713 mGy-cm (median: 627)
6. Chest with contrast—mean DLP 711 mGy-cm (median: 486)
7. Cervical spine without contrast—mean DLP 706 mGy-cm (median: 555)
8. Neck with contrast—mean DLP 620 mGy-cm (median: 513)
9. Chest pulmonary arteries with contrast—mean DLP 561 mGy-cm (median: 459)
10. Chest without contrast—mean DLP 448 mGy-cm (median: 367)

Escalon and colleagues additionally found that, compared with academic and community hospitals, lower mean DLPs were observed at children’s hospitals, multispecialty clinics and freestanding centers.

In their study discussion, the authors note several limitations inherent to their study’s design, including the voluntary nature of the ACR registry. This, they point out, could prompt facilities with unflattering data to steer clear.

They also take into account DLP’s shortcomings as a measurement of dose, including its reliance on standardized phantoms. Such reliance, they allow, constricts DLP’s capacity for accurately reflecting individual, patient-effective dose.

“However, DLP can be used to directly compare scanner output and has been used as a measure of radiation exposure in several studies,” they write. “Additionally, use of DLP has precedent in the literature, as a proxy for overall dose levels for the purposes of comparing relative output between sites, scanners, and individual examinations.”

Stressing that their observations are preliminary, the authors express hope that the dissemination of their findings as baseline data will help individual facilities “understand their position relative to national averages, and will help us move toward defining and optimizing appropriate limits on radiation exposure for individual examinations.”

“With a dual focus on both diagnostic quality and radiation exposure,” they conclude, “radiology facilities and individual practicing radiologists can help reverse or stabilize current trends of increasing medical radiation exposure.”

Click here to explore the full study.