The AAPM Statement on Radiation Dose from Computed Tomography, in response to the Brenner and Hall NEJM article published Nov 29, 2007.
November 30, 2007
A recent article by Drs. David Brenner and Eric Hall in the New England Journal of Medicine1 has suggested that the radiation dose from CT scans is a cause for concern, and may be responsible for a small percentage of cancer deaths in the United States. While the conclusions of the Brenner article have been portrayed by some as conclusive, in reality the scientific community remains divided in regards to the radiation dose effects of CT. The AAPM is an organization of 6700 Medical Physicists, and radiation dosimetry in CT and other sources of x-ray exposure is the core expertise of the vast majority of our members. Dr. Brenner’s article correctly points out that the use of CT is increasing at an exponential rate, and that CT should not be used for medical indications that are not warranted or serious. The AAPM adamantly concurs with these observations, and has long advocated that CT should be used judiciously and only when medically indicated. For example, the AAPM policy paper on CT screening has recommended against “CT screening” for many years.
However, the medical information that is derived from appropriate diagnostic CT scans literally and emphatically saves the lives of thousands of Americans on a daily basis, for patients who have experienced severe trauma such as in an automobile accident, for patients who have blood clots in their lungs (pulmonary embolisms), and for a vast array of other medical indications. The Brenner article illuminates many issues of importance in regards to CT, but the CT experts in the AAPM feel that much of the message of this article may be misconstrued or misunderstood by the press or by the public who may not be experts in CT.
The assumptions used in Brenner’s article remain controversial even among experts in radiation biology. The underlying data in Brenner’s calculations are based upon a number of controversial assumptions – for example, the risk coefficients used by Brenner are derived from studies of the Japanese citizens who were exposed to large amounts of radiation during the A-bomb attacks of Hiroshima and Nagasaki 62 years ago. These individuals were exposed head to toe to a mixture of x-ray and particulate radiations, whereas CT examinations involve only x-rays over a small fraction of the patient’s body, for example the head or the abdomen. In addition, the majority of the A-bomb survivors experienced radiation doses many times that of modern computed tomography, and the mathematical methods (called extrapolation) used in scaling the very high radiation exposures levels down to the much lower exposure levels typical of partial body CT examinations remains very controversial. Indeed, the underlying risk data used by Brenner and Hall used the most extreme mathematical assumptions – those which would predict the most harm from CT. The data from the Japanese bomb survivors is also consistent with more moderate risk assumptions.
Another significant flaw in the assumptions used by Brenner and Hall is that the radiation risks derived from the Japanese studies are applicable to the patients receiving CT in the US in 2007. Patients who require medically indicated CT scans, in the broad brush of generalities, are sicker than most Americans and thus have greater health risks and are far likelier to benefit from the diagnostic information that the CT examination provides to doctors involved in their patient care. The population of patients undergoing CT is also significantly older than the normal population, and although Brenner and Hall corrected for age using data derived from the 1945 Japanese population, they did not correct for the many underlying confounding age dependent variables that differ between this population and older Americans, such as the incidence of obesity and diabetes.
The bottom line is that patients and parents or loved ones of patients who have had CT scans, or are slated to have CT examinations in the next days and weeks, should discuss with their physicians not only the radiation risks of the CT examination, but the risks of not having the diagnostic information that CT provides. Before the invention of CT (in 1972), exploratory surgery was common practice. CT and other imaging procedures have virtually eliminated the need for exploratory surgery, since these technologies allow doctors to peer inside the patient without the use of a scalpel. Nobody wants to go back to the days of exploratory surgery, which has a number of significant risks including that of bleeding to death, infection, or debilitating nerve damage. While the AAPM strongly endorses the Brenner and Hall concerns that CT should be used only when medically necessary, we have the steadfast belief that the medical information gained by medically indicated CT studies leads to better medical decisions, better patient care, and a significant improvement in human health.
David Brenner and Eric Hall are esteemed scientists and respected experts in radiation risk, and in no way is this synopsis meant to impeach or undermine their impressive credentials. Nevertheless, while it is assumed by many in the lay press that science should be definitive and consistent, the consequences of the radiation exposure from CT procedures remain subject to interpretation of the sparse data that are available. It is emphasized that the conclusions of the Brenner article are based on statistics and many statistical assumptions, they are not based on the actual observation of somebody dying from having a CT scan.
If questions among patients or referring physicians remain in regards to the radiation risks of CT, they are encouraged to contact medical physicists who are employed by their local medical center or academic radiology department.
1. DJ Brenner and EJ Hall, Computed Tomography – An Increasing source of radiation exposure, NEJM 327:22, 2277 (2007).