Exposure of Patients to Ionizing Radiation. What are the Risks?

Over the last 35 years, exposure of the general public to ionizing radiation for medical purposes has increased remarkably. Primarily due to frequent use of Computerized Axial Tomography (CT) in the detection and monitoring of many pathologies, in both intpatients,¹ and outpatients,² dose delivery to the population has increased from 3.6 mSev to 6.2 mSev³ from 1980 to 2006. In 1980, 20% of total patient exposure was medically related whereas, in 2006, 51% of the total was medically related (4). This increase, amplified and distorted by the electronic and print media, has led to unease and fear among patients who have clear need for imaging studies. I would like, therefore to review the appropriate literature, indicate the degree of trust we may or may not place in these findings, and offer conclusions we may accurately convey to our patients.

First: The use of techniques involving ionizing radiation has proliferated, particularly with respect to the use of CT. This increase is ubiquitous, occuring in all populations and methods of reimbursement.⁵

Second: There is little to no useful data dealing with exposure of human populations to low level radiation. The data previously used to derive risk rates from imaging procedures has been extrapolated from higher doses (BEIR V11 data)) using controversial techniques.⁶

Third: We must separate relative risk from absolute risk, the key to allowing an understandable conclusion. Relative risk increase is clearly documented, being dose and age dependent as demonstrated in multiple studies.^7,8 Incremental absolute risk (benefit to risk ratio) is dependent on the underlying pathology⁹ and has been much more difficult to measure, requiring accurate data acquisistion and analysis using meticulous technique. Happily, the Welsh have done it. Pearce, et al, showed,¹⁰ in a population with no pre-existing neoplasms, that exposure to 3 thorax / abdomen scans or 5–6 head scans increased relative risk of lymphoma / leukemia by three times. The cumulative absolute risks (experimentally measured and not extrapolated from high dose patients) however, remained small: one excess case of leukemia, one excess case of brain neoplasm per ten thousand head scans (.01–.02%). This is the most accurate measurement that we have to date and appears quite valid.

Fourth: Any discussion of “risk” involving imaging procedures must include a more significant issue: incorrect diagnosis. Literature over the last 40 years has addressed this topic, showing marked variation in error rates. Berlin's article clearly summarizes the topic and notes an apparently irreducible error rate of 3–4 %.¹¹ This rate is stable over multiple studies and also applies to children.¹²

Conclusions: The recent discussions in the media sensationalizing radiation dosage morbidity and mortality require a balanced analysis in order to evaluate our approach to imaging for our patients. The best references for an inquiring patient or health care worker are #6 & #10. They are clear, as scientifically accurate as possible, and cover the topic evenly. The full text of #6 is available online without charge. We must also accept that unnecessary radiation can have deleterious effects without measurable benefit and that a major risk is the apparent irreducible error rate of 3–4%.