Variable doses of radiation raise safety concerns for CT procedures

"In day-to-day clinical practice, we found significant variation in the radiation doses for the same type of computed tomography procedures within institutions and across institutions," said lead investigator Rebecca Smith-Bindman, MD, a professor of radiology at UCSF. "Our results highlight the need for greater standardization because this is a medical safety issue."

Computed tomography imaging, known as CT, is a diagnostic procedure that uses special x-ray equipment to obtain cross-sectional pictures of the body that provide detailed images of organs, bones, and other tissues. CT is associated with higher radiation exposure than conventional x-rays, yet radiation dosages that patients receive from the newer CT scanners have gone largely unregulated, according to Smith-Bindman, who also is a UCSF professor of epidemiology and biostatistics and obstetrics, gynecology and reproductive sciences.

"Our study provides some initial data documenting the doses that patients receive when they undergo actual CT examinations and this is different than the doses when phantoms-- sophisticated plastic models typically used to quantify CT scanner dose-- are used. We believe documenting the actual doses that patients are exposed to is the first step to reducing those doses and any attendant risk," she emphasized.

Since 1980 the yearly number of CT exams has increased from about 3 million to 70 million CT procedures. The technology has changed dramatically over that time, improving the quality of imaging, and increasing the clinical questions that could be answered using CT, therefore leading to improvements in patient care, according to Smith-Bindman.

One of the improvements in CT technology has been the dramatic decline in the speed it takes to complete a CT examination. "However, this has been a double-edged sword," she said. "Because the images can be obtained so quickly, it has been very tempting to do multiphase studies -- one CT done during the arterial phase of the cardiac cycle, one CT done during the venous phase, and one done after a short delay. This increases the information that we can get from the CT procedure, but increases the radiation dose by a factor of three."

The research team focused on estimating the radiation exposure associated with the 11 most common types of CT procedures in clinical practice in the U.S. and the potential cancer risk associated with each type.

The study is the first large-scale examination of radiation doses given to patients who underwent a range of different types of CT procedures. The researchers reviewed the procedures performed on 1,119 patients at four San Francisco Bay Area institutions over five months. They evaluated procedures in three anatomic areas: head and neck, chest, and abdomen and pelvis.

Study findings appear in the December 14, 2009, issue of the Archives of Internal Medicine.

Study results showed that doses were in general higher than typically reported and that there was a mean 13-fold variation between the highest and lowest radiation dose for each type of CT procedure, which means "If a physician sent a patient for a particular CT procedure, the dose that patient would have received varied by this much," Smith-Bindman explained. As expected, the researchers' estimate of the number of CT scans that would result in a cancer varied widely by sex, age, and type of scan. They estimated that it would take fewer CT scans to result in a cancer among women compared with men.

They estimated, for example, that 1 in 270 women who underwent a CT coronary angiogram at age 40 years will develop cancer from the procedure, compared to 1 in 600 men. For a routine head CT scan at the same age, the estimated risk is 1 in 8,100 for women and 1 in 11,080 for men. In patients around age 20 years of age, the risks were approximately twice as high. "The risk associated with obtaining a CT is routinely quoted as around 1 in 1,000 patients who undergo CT will get cancer. In our study, the risk of getting cancer in certain groups of patients for certain kinds of scans was as high as 1 in 80," noted Smith-Bindman.

The research team used "effective dose" to quantify the radiation exposure associated with each CT scan because it is one of the most frequently reported measurements. Then to translate the variation of doses into cancer risks, the team utilized data from a National Academy of Science National Research Council report that assessed the health risks from exposure to ionizing radiation. The report was published in 2006 under the title "Biological Effects of Ionizing Radiation VII Phase 2."

Depending on the location and technical parameters of the individual CT scanner, the effective dose received by a patient could substantially exceed the median, according to study results. "This is particularly important because the threshold for using CT has declined, and CT is increasingly being used among healthy individuals, in whom the risk of potential carcinogenesis from CT could outweigh its diagnostic value," the research team noted in the "Archives" article.

Compared to other imaging procedures, the median effective dose delivered through a single CT scan was as high as 74 mammograms or 442 chest x-rays, according to Smith-Bindman.

The researchers identified three key practices necessary to improve the safety of CT procedures and the associated radiation doses:

• Reduction of unnecessary studies and studies thought unlikely to influence clinical decisions.
• Standardization and utilization of low-dose and lower-dose protocols for every CT scanner.
• Standardization of radiation doses across patients and facilities through federal legislation and FDA oversight stipulating how CTs are to be safely performed.

Smith-Bindman emphasized that further investigations into CT safety are a must. "It is imperative, particularly given these results, that we start collecting radiation dose data at the individual patient level. Our results point toward the need to start collecting data on what actually happens in clinical practice and then to establish the appropriate standards."

In addition to publication in the Archives of Internal Medicine, Smith-Bindman will present the research findings at the 2010 spring meeting of the National Council of Radiation Protection in Washington, DC.

Study co-authors are Jafi Lipson, MD, and Robert Gould, ScD, UCSF Department of Radiology and Biomedical Imaging; Kwang Pyo Kim, PhD, Kyung Hee University in Korea; Mahadevappa Mahesh, MS, PhD, Johns Hopkins School of Medicine; Amy Berrington de Gonzalez, DPhil, National Cancer Institute; and Diana L. Miglioretti, PhD, Group Health Research Institute, Seattle.

The study was funded by the NIH National Institute of Biomedical Imaging and BioEngineering, NIH National Cancer Institute, and the UCSF School of Medicine.