Study: Diagnostic X-ray Exposure Risk Low

Findings Link Diagnostic Imaging to 1% of Cancers in U.S. and 3% in Japan
 

Jan. 29, 2004 -- Roughly one in 100 cancers in the U.S. are caused by exposure to medical X-rays, according to the most detailed study of radiation risk associated with diagnostic imaging ever conducted.

Conventional X-rays and imaging techniques such as computed tomography (CT) are the largest man-made source of radiation exposure. While the benefits of these diagnostic tools are well established, the risks have been harder to quantify.

"There have been relatively few studies looking at this issue," University of Oxford research fellow Amy Berrington de Gonzalez, PhD, tells WebMD. "It is generally accepted that their use is associated with a small increase in cancer risk, but the extent of that risk has not been clear."

Estimates of Cancer Risks

Berrington de Gonzalez and colleague Sarah Darby, estimated the likely extent of cancer risk for 15 industrialized countries, based on the annual number of diagnostic X-rays given each year.

People in Japan had the highest risk, because they underwent more diagnostic imaging procedures than people in any of the other studied countries -- 1,477 for every 1,000 people. That compares with:

* 1,254 X-rays annually per 1,000 people in Germany
* 962 X-rays annually per 1,000 people in the U.S.
* 565 X-rays annually per 1,000 in people Australia

The researchers concluded that almost 7,600 cancers diagnosed each year in Japan, or 3.2% of the total cancer cases, are linked to diagnostic imaging procedures.

The people in the U.K. had the lowest risk, with just 489 X-rays conducted each year for every 1,000 people. The researchers estimate that the lifetime risks attributed to radiation exposure from diagnostic X-rays is roughly 0.6% in the U.K. That translates to roughly 700 of the 124,000 cancer cases identified in the country each year. The findings are published in the Jan. 31 issue of the journal The Lancet.

They also estimate that in the U.S. the lifetime risk of cancer caused by radiation exposure from X-rays is approximately 0.9%. That is almost double the risk suggested in one of the only other studies ever to examine the issue, published in 1981. This lifetime risk accounts for about 5,700 cancers per year.

Risk Estimate Challenged

In an accompanying editorial, University of Munich radiologist Peter Herzog and oncologist Christina T. Rieger write that the researchers did not assess the indications or benefits achieved for patients in X-ray examinations. Benefits include the earlier detection of cancers by radiological examinations and the possibility of early treatment, which probably allows more cure of cancers than radiological exposure is able to cause.

Herzog tells WebMD the study used cancer rate information from Japanese bomb survivors as a model to study the risks of cancer from diagnostic X-ray exposure. They had very different exposures, most notably those associated with ingesting contaminated food, water, and dust.

"The problem is that these bomb survivor data are really the only studies examining radiation exposures in a large population," he says. "But we believe that relying on them leads to a very great overestimation of the real risk."

Even though he believes the risk is probably minimal, Herzog says clinicians should take steps to limit their patients' radiation exposure by reducing unnecessary X-rays and CTs.

"Up to 30% of chest X-rays may not be indicated; unnecessary CT examinations can lengthen hospital stays as well as causing radiation exposure," Herzog and Rieger wrote in their editorial. "In everyday practice, those ordering radiological procedures should think carefully about the benefit for and the risk to their patients for each examination."

SOURCES: Berrington de Gonzalez et al., The Lancet, Jan 31, 2004; vol. 263: pp - 345-351. Amy Berrington de Gonzalez, PhD, research fellow, University of Oxford, Oxford, U.K. Peter Herzog, MD, radiologist, University of Munich. Christina Rieger, MD, oncologist, University of Munich.