CT is one of the most transformative advances in medicine in the past 100 years. It is estimated to have a substantial impact on the speed of diagnosis in 50% of cases and, in a high proportion, the result is decisive in outcome. Approximately 15 per 1000 people undergo a CT scan annually in Australia.
Radiation is possibly the longest-studied carcinogen, with reports dating back to the early 1900s. Radiation risks are divided into so-called deterministic (acutely after large doses of radiation e.g. radiotherapy) or stochastic (arising in the longer term, usually decades after a significant exposure). The stochastic effects are applicable to CT, chief among them is cancer.
There are expressed concerns about the safety of CT scans though these are often not consistent with data from higher quality studies nor the consensus opinions of authoritative bodies.
Quantification and appropriate dose
Scanner software provides an accurate dose estimate on a per patient basis, expressed in mili Gray per centimetre – this appears in the dose card image on the image viewer (PACS) as the Dose Length Product (DLP). In the radiologist’s report, this may have been converted to mili Sieverts (mSv), a measure weighted to the particular radiation sensitivity of the region scanned. For example, a DLP of 200 mGy.cm for the head yields an effective dose of 1 mSv, though for the chest this same DLP translates to 3 mSv, reflecting the differences in tissue sensitivity of these two regions.
The choice of the dose is the attending radiologist’s, in consultation with the radiographer, when needed. Firstly, it is incumbent on radiologists to ensure the scan is warranted and that CT is the best imaging modality to answer the clinical question.
To enhance outcomes, radiologists are periodically required to submit their radiation dose records to the Australian Radiation Protection and Nuclear Safety Agency (ARPANSA) and are provided with updated diagnostic reference levels to ensure they are within reasonable range of their peers.
What is low-dose CT and who gets one?
While there is no broadly accepted definition of what constitutes a low-dose CT, it is commonly reserved for scans less than 1mSv. Thresholds will undoubtedly change, or terminology shift (e.g. the recent use of ‘ultra’ or ‘hyper’ low dose) as technology continues to improve.
Importantly, a low-dose CT is often inappropriate (e.g. characterising a renal mass may require a standard dose of 10 mSv though excluding a renal stone just 1 mSv). Other indications for low-dose CT include lung cancer screening, paranasal sinus imaging, virtual colonoscopy and coronary calcium scoring.
Doses do vary and will usually reflect the particular requirements for the clinical question or result from patient-specific factors. For example, the dose used will often be higher for an obese patient (though curiously not necessarily the biological effect because of the radioprotective effect of subcutaneous fat).
When do I request an alternative test?
Each imaging modality (e.g. MRI, ultrasound) has strengths and weaknesses with regards cost, availability, comfort and fitness for purpose. For broad recommendations, we refer to the published diagnostic imaging pathways and, if there is persisting uncertainty, a radiologist should be consulted. A user-friendly pathway (developed in WA and now adopted by the UK National Institute for Care and Health Excellence) can be found at: https://radiologyacrossborders.org/diagnostic_imaging_pathways/
CT scans and cancer risk – the evidence
CT scans cause acute DNA damage (double strand breaks), however, this process is also occurring naturally and continuously and typically managed by the body’s regulatory system before generating cancer. The highest (and only direct) level of evidence of radiation-induced carcinogenesis is derived from survivors of the atomic bomb blasts in Hiroshima and Nagasaki. These studies show a linear dose-risk relationship for individuals receiving 100mSv or above.
Some patients will undergo numerous CT scans during their lifetime, though with current and foreseeable applications and technological advancements, this will unlikely exceed an accumulation of 100mSv or more. This excludes some patients with a significantly shortened life expectancy (e.g. those with poorer prognosis cancers).
Some recent indirect evidence suggests CT scans cause cancer. However, this evidence is significantly flawed including high uncertainty about the doses, lack of detail about scan indication or past medical history, and no consideration of doses from other imaging tests. Hence the potential for reversed causality bias is high, as it is in most published radiation risk science.
The preponderance of data, including that from other recent studies, taking such factors into account (though there may be other limitations), has not shown an excess in cancer risk from CT scans, though these studies have not proven the risk is zero.
The effect (if any) of CT scans is simply too small to be registerable with the data available to date. Several studies with improved designs are underway.
In children, some tissues are more sensitive, some have the same sensitivity and others less sensitive to radiation than adult tissues. Thus far the evidence is too weak to establish an association between CT radiation and cancer in children.
Nevertheless (and fortuitously), doses required for diagnostic acceptability in children are typically smaller than those in adults. In addition, there has been extraordinary effort made by radiological colleges in recent years to increase awareness and promote particular attention to minimising doses in children.
Key messages
- Radiation doses from CT scans are low (and getting lower) and the risk is small to none.
- Radiologists are bound to strict radiation hygiene with doses tailored to purpose and no more.
- A radiologist considers the appropriateness of every scan prescribing dose in accordance with patients’ best interests.
Note – radiation refers to ionising radiation
Author competing interests – Dr Murray has a commercial interest in a provider of CT scan services