Over Exposure

30 March 2008 (Last Updated March 30th, 2008 18:30)

Medical imaging has proved to be such an invaluable diagnostic tool that its use has grown dramatically in recent years. However, the benefits of imaging technology must be carefully balanced against the risks associated with radiation exposure, says Dr Eugenio Picano.

Over Exposure
The indiscriminate use of CT and X-ray scanning gambles with patients’ genetic health.

Many imaging technologies offer welcome diagnostic benefits. As a result, in the US, diagnostic imaging services, reimbursed under Medicare's physicians' fee schedule, grew more rapidly than any other type of physician service between 1999 and 2003. However, there is concern over the escalating costs and risks associated with some of these imaging technologies, especially as at least one in three medical imaging tests is inappropriate.

An appropriate imaging study is one where the expected benefits of the information gained exceeds the expected negative consequences sufficiently and where the procedure is considered to be acceptable care and a reasonable approach for the indication. In imaging, acute risks are contrast, procedure and stress related – anaphylaxis during contrast injection, myocardial infarction durin stress, and contrast-induced nephropathy, for example. The order of magnitude of all acute risks is one in 100.

Compared with these readily apparent acute adverse events, the radiation risk posed by ionising testing is difficult to evaluate. The risk is invisible, long term and cumulative – every examination compounds the dose and the risk. It is significantly modulated by polymorphisms of genes involved in DNA damage and repair (such as the BRCA1-BRCA2 mutation). The target molecule is DNA, and target cells are actively dividing somatic cells (such as blood stem cells) for cancer effects, embryo cells for teratogenic effects and germ cells for adverse hereditary effects.

Long-term adverse radiation effects are a major and rarely acknowledged concern with X-ray imaging, but they must be factored into daily decision-making to assess the risk-benefit balance properly, not just for patients but for health workers, too.


It is important to know the radiological doses involved in medical imaging, because the radiological dose is directly and linearly related to risk. Radiological doses can be expressed as dose equivalents of multiple chest X-rays, as suggested by the European Commission's medical imaging referral guidelines. The doses associated with common medical imaging examinations are reported in the original data of European Commission guidelines 2001 and the Italian health ministry guidelines of

Figure 1 shows the corresponding risk of cancer, following the latest estimates of the Biological Effects of Ionizing Radiation (BEIR) 2006 report, which provides a framework for estimating the lifetime attributable risk of cancer incidence from radiation exposure using the most current data on the health effects of radiation. It is one order of magnitude lower than most acute risks. For example, one multi-detector body CT scan (750 chest X-rays) gives an extra risk over a lifetime of one
cancer for every 750 patients. The average risk varies as a function of both gender and age. For women, the risks of developing cancer after exposure to radiation are 37.5% higher than they are for men.

According to these estimates, the risk of cancer for a 15mSv exposure (corresponding to a multi-detector body CT scan) is one in 750 for an adult (20-50 years range), 1 in 500 for a woman, and 1 in 1,500 for an elderly patient. A 64-slice CT coronary angiography without tube current modulation is associated with a cancer risk ranging from one in 143 for a 20-year-old woman to one in 5,017 in an 80-year-old man for a scan performed with tube current modulation.

Small individual risks multiplied by widespread testing (60 million CTs are performed in the US annually) imply a significant population risk. In developed countries, exposure to radiation from medical ionising tests results in a mean effective dose per year per head ranging from 100 chest X-rays (in Germany, 1997) to 160 (USA, 2006 – equivalent to more than a year of natural background radiation). In 2004 Barrington and Darby estimated that 0.6% (in the UK) to 3.2% (in Japan) of
cancers could be caused by diagnostic X-rays.

The attributable cancer risk from diagnostic X-rays was 0.9% for the US and 1.5% for Germany. In 1991–96, the mean exposure for the US citizen was 0.6mSv per head per year from X-rays. Since then, the number of medical exposures has risen substantially, especially for CT, interventional cardiology and nuclear medicine, accounting for more than 90% of cumulative patient exposure. In 2006, the estimated exposure (from radiology and nuclear medicine) reached an unprecedented 3.2mSv per
head, per year, more than six times higher than the estimate used by Barrington. The attributable cancer risk will rise accordingly – at least 5% of cancers could result from diagnostic radiation.


In a modern, high-tech hospital, patients receive an average cumulative dose – from appropriate and, in 30% of cases, inappropriate examinations – of up to 100mSv (about 5,000 chest X-rays) in one case in four, leading to an estimated additional risk of about one (fatal or non-fatal) cancer in every 100 exposed subjects.

"At least 5% of cancers could result from diagnostic radiation."

The risk model of Biological Effects of Ionising Radiation Committee VII for exposure to low-level radiation predicts that about one in 100 people are likely to develop solid cancer or leukemia from an exposure of 100mSv above background. This exposure was reached by about one fourth of our study population.

About 42 additional people in the same group would be expected to develop solid cancer or leukemia from other causes. Roughly half of these cancers would result in death. It must be acknowledged that current risk estimates suffer from some degree of approximation and uncertainties – it can be two or three times higher, or lower, than current estimates.


In recent years, it has been clearly shown that prescribing and practicing doctors are unaware of, or substantially underestimate, by between ten and 1,000 times, the dose of commonly employed radiological examinations and the corresponding risk to patients. Among US radiologists, 5% thought that an abdominal CT scan dose (about 500 chest radiographs) was less than one chest radiograph, while 56% estimated the dose to be between one and ten chest radiographs. Clearly it is difficult to carry out a proper risk-benefit assessment if long-term risk is not taken into account. Patients have little idea of the dose received during common examinations, such as a CT or a cardiac stress scintigraphy, and the consequent risks they face.

Common sense, deontological code, patients' rights, medical imaging guidelines and Euratom law all recommend the justified, optimised, responsible and informed use of testing with ionising radiation. An informed consent form should be filled out for high-radiation tests detailing the type of examination to be conducted, its dose equivalent in terms of chest X-rays and the risk of cancer, expressed, for example, as the number of extra cases in the exposed population. A more explicit and transparent consent form would help reduce pressure from patients for redundant and often useless examinations.


Occupational doses in interventional procedures guided by fluoroscopy are the highest doses registered among medical staff using X-rays. Cardiologists and interventional radiologists experience radiation exposure with the patient, as they are close to the radiological source, whereas diagnostic radiologists are generally shielded from radiation exposure.

For highly exposed personnel with 5mSv of yearly exposures, the lifetime extra-risk for cancer after 20 years of professional life is in the range of one to 100. Interventional cardiologists also show an increased frequency of micronuclei in circulating lymphocytes, which are a marker of significant somatic DNA damage, an intermediate endpoint of carcinogenesis and a long-term predictor of cancer. This radiation level remains well below the dose limit of occupational exposure, consisting of
an effective dose of 20mSv in a year averaged over five years. Still, it is neither negligible nor harmless.

"Cardiologists pay little heed to monthly or cumulative reports of radiation exposure"

Some of the exposure and damage to professionally exposed staff is unavoidable, due to the high radiation burden of the imaging procedure and the close presence of the operator near the radiation source. However, some exposure is avoidable, specifically, exposure from the inappropriate use of imaging and exposure resulting from failure to follow the as 'low as reasonably achievable' (ALARA) principle in the cath lab.

Cardiologists pay little heed to monthly or cumulative reports of radiation exposure, and while some institutions require a course in radiation safety prior to issuing radiation dosimetry badges, they are the exception. There is often a macho disregard for radiation protection.

If radiation machismo stays unchecked, avoidable exposure of patients and doctors will remain. An intensive training programme in radiation protection can reduce occupational doses by a factor of ten. Recent studies confirm that simple, effective protection measures (such as a lead curtain, protection glasses and thyroid collars) are not used by the majority of interventional cardiologists.

Strzelczyk et al have provided an excellent brief review on managing radiation exposure in the interventional fluoroscopic environment. They recommend that healthcare workers use of wrap-around two-piece lead aprons and thyroid shields for protection. They also recommend leaded eye glasses with side-shielding and flexible, disposable radiation protective drapes. Clearly more research is needed into occupational exposure.