How safe are CT scans in children?


Share on facebook
Share on twitter
Share on linkedin
Share on whatsapp

“I am worried that they may have developed pneumonia, which is a bacterial infection in the lungs. What I want to do next is get a chest x-ray to see where it is exactly, and if there are any other issues in the lungs, like a collection of fluid,” you say to the parents of the two-year-old sleeping in the bed, with an IV drip connected for antibiotics and supplemental oxygen already commenced.

They look at each other, and then the father replies, “But doctor, they were born 12 weeks premature, and have already had so many x-rays… do they really need another one? Aren’t too many x-rays bad for the body?”

You pause and think.

As paediatric emergency physicians, we strive to provide appropriate, timely and safe care. We rely heavily on history taking, clinical examination and gestalt to reach a diagnosis and management plan. We try our hardest to avoid unnecessary investigations which may cause distress and pain. Radiological interventions fall under that category – although they can provide valuable information, have you ever thought to yourself, what is the true radiation risk?

When it comes to radiological interventions, the aim is to minimise patient exposure to ionising radiation. Radiation is simply energy that moves through space. It has many forms including radio waves, microwaves, infrared radiation and visible light. These are all non-ionising.

So, what are x-rays? X-rays are a type of ionising radiation generated by electricity used in producing plain x-ray films, fluoroscopic and CT images. As these x-ray waves pass through the body, different materials (bone, tissue, air, fluid) absorb the rays in various amounts to create the images we see and interpret. Ionising radiation removes electrons, leaving behind positive ions (free radicals) from molecules of air, water and living tissue (more on free radicals later).

Radiation is typically measured in:

  • Radioactivity – how many radioactive atoms decay per second.
  • Radiation dose – the amount of radiation absorbed by the body.

Units of measurement are reported in Gray (Gy) in the United States and Sieverts (Sv) in the rest of the world. The natural background radiation differs in every country due to weather conditions, cosmic rays reaching the earth, the type of soil and rock and the vegetation.

The global average background radiation is 2.4mSv/year. It’s a little higher in the US at 3.1mSv/year, a little lower in Australia at 1.5mSV/year and just above average in the UK at 2.7mSv/year.

What does this mean? Well known comparators, Chernobyl and Pripyat, had background radiation of approximately 9mSv/year for the 20 years after the nuclear disaster, much higher than the current global average. For those physics nerds reading this, The United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) has published over 25 articles looking at the health impacts of radiation.

How does medical ionising radiation exposure compare?

Mettler et al nicely summarised the adult radiation doses of a range of plain x-ray films, fluoroscopic imaging, CT images and nuclear medicine interventions and imaging in their 2008 publication. Although this article is over 10 years old, there hasn’t been a need to republish the data as it remains pretty accurate. For the paediatric population, Gibbo (2017) nicely summarised the paediatric radiation dose ranges (by age group) for plain X-ray films, fluoroscopic imaging, CT imaging and nuclear medicine interventions and imaging.

Paediatric radiation doses associated with different x-rays are summarised below:

Based on background radiation activity in Australia, which is 1.5mSv per year (approx. 0.004109mSv/day), we can work out that for a five-year-old:

  • An AP chest x-ray is equivalent to 3 days’ worth of background radiation;
  • An AP abdominal x-ray is equivalent to 9 days’ worth of background radiation;
  • A CT brain is equivalent to 297days’ worth of background radiation;
  • A CT abdomen/pelvis is equivalent to 990 days’ worth (approx. 2.7 years) worth of background radiation.

What about those free radicals?

The severity of any effect is dependent on the total amount of exposure to radiation within a given time, known as the radiation dose. There is a direct damaging effect of ionising radiation on DNA in living material. In addition, once electrons are removed, the positive ions, or free radicals, can also damage DNA. If a person is exposed to very high levels of radiation for a prolonged period of time, the accumulated dose is obviously much higher. Once DNA is damaged, there are three outcomes:

  • Cell death (occurs with very high doses).
  • Imperfect cell repair (the most common outcome), which can lead to the development of cancer.
  • Perfect cell repair (least common outcome).

The UNSCEAR, based on studies of the populations exposed to radiation following nuclear incidents in Japan and Chernobyl, have been able to provide some guidance on the dose of radiation and the subsequent effects.

Graphic showing dosage of radiation that leads to risk

Age is a huge factor. Paediatric and (especially) neonatal bodies are growing rapidly, and thus have more susceptible, rapidly dividing cells. Their longer lifespan means that earlier and more repeated exposure to ionising radiation gives a double load of increasing the risk of developing cancer over time.

The studies

Epidemiological studies looking at the incidence of cancer following high exposure to ionising radiation have been the basis for risk modelling in studies of medical radiation exposure. Whilst there are not many well-conducted studies in the paediatric population, Berrington de Gonzalez et al (2004) published the results of a large observational study across 14 countries looking at the risk that an individual will develop cancer based on the annual x-ray frequency per country.  The models for cumulative risk took into consideration the background cancer incidence in each country, the x-ray frequency based on the type and the estimated organ-specific radiation dose delivered by each type of x-ray. In the UK, a rate of 489 annual x-rays per 1000 equates to 700 cases of cancer per year (in a population in 2021 of 68 million) while in Australia, a rate of 565 annual x-rays per 1000 equates to 431 cases of cancer per year (in a population of 25 million).

Given the lack of data around paediatric x-rays and cancer risk, it is difficult to provide parents with clear and consistent information. Bibbo (2018) used data from x-rays, CT, fluoroscopy and nuclear medicine scans in children and young adults aged 0 to 18 to provide a summary of the effective dose of radiation per scan and radiation-induced cancer risk. Minimum, maximum and average doses were calculated. Reassuringly, the lifetime risk of radiation-induced cancers ranged from negligible to low. But… these results are the “standardised” risks, and the actual risk to each patient may be different – the dose absorbed depends on the examination performed but also patient age and body size (height and weight).

What about neonates?

There are several studies confirming that premature babies are exposed to more radiation than term babies admitted to the NICU. Interestingly, this trend continues in the first year following discharge from the hospital. following discharge from the hospital. Hogan et al (2018) published a seven-year retrospective cohort study calculating the odds of premature and term infants exceeding the recommended radiation exposure threshold of 1mSv in the first year after discharge from hospital. They found that premature infants had 2.25 times greater odds of crossing the threshold compared with term infants. Even following adjustments to consider the confounding variable of paediatric complex chronic conditions, premature infants still had 1.58 times greater odds.

And CT?

We know CT delivers a higher radiation dose than x-rays, and as an imaging modality, is being increasingly used in developed countries. The elephant in the room is concern about the risk of developing leukaemia (following radiation absorption into red bone marrow) or solid brain tumour (following radiation absorption into brain tissue). The question is, how big is the risk? A retrospective cohort study from Pearce et al (2012) looked at NHS patients under the age of 22 who received a CT scan. The authors found a positive, but statistically non-significant, association between the radiation dose from CT scans and leukaemia (p = 0.0097) and a statistically significant risk of subsequent brain tumours (p<0.0001). The relative risk of both cancers increased with increasing cumulative doses. What does this mean for our patients? The incidence of both of these cancers is low; in the 10 years after the first CT scan in children under 10, there was estimated to be only one extra case of leukaemia and one extra brain tumour per 10,000 head CT scans.

In line with other studies, Berrington de Gonzalez et al (2016) found a statistically non-significant positive excess relative risk (ERR) of developing leukaemia and a statistically significant positive ERR of developing a brain tumour. After the data underwent sensitivity analyses and excluded participants who had underlying cancer predisposing conditions, the risk was reduced for both groups, although remained statistically significant for brain tumours.

There are plenty more studies looking at the ERR of leukaemia and brain tumours after CT. Perhaps one last notable one is by Muelepas et al (2019). This Dutch team looked back on data of children under 18 who received one or more CT scans between 1979 and 2012. In addition to CT scan type and estimated radiation dose and organ absorption, the authors also obtained cancer incidence with a two year lag time for leukaemia and a five year lag time for solid brain tumours. The authors found that the mean cumulative bone marrow doses were 9.5 mGy and, of note, leukaemia risk was not associated with cumulative bone marrow dose (ERR/100mGy: 0.21, 95% CI, p = 0.68). However, brain tumours were a different story. The mean cumulative brain dose was 38.5 mGy with a statistically significantly increased risk for malignant and non-malignant brain tumours combined (ERR/100 mGy: 0.86, 95% CI, P = .002). Despite correcting for biases such as household income, or increasing/decreasing the lag time, the results remained similar with a non-statistically significant increased risk of leukaemia and a statistically significant increased risk of brain tumours.

So, what about our toddler with pneumonia?

Knowing what you know now, how would you explain the radiation risk to parents? Can you alleviate some of their fears given their child’s medical history? Oikarinen et al (2019) surveyed parents to assess communication received about radiographic examinations, asking parents what would they like to know when their child requires medical x-ray exposure. Most parents (83%) received adequate information about the purpose of the examination but only 7% received information about the radiation dose. Approximately a quarter of parents (23%) received information about other imaging modalities.  The majority of parents not only expected information on the purpose of the imaging (95%) and options for alternative imaging (78%) but also wanted information on radiation dose for the imaging or procedure their child was to have (88%).

Taking into consideration a child’s medical history and using the standardised estimated dose absorptions tables calculated, you can pretty confidently guess what the mean cumulative radiation dose has been over their lifetime.  If you really want to be specific, you can calculate the risk using an x-ray risk calculator like this one.

Practising evidence-based medicine and establishing a holistic doctor-patient relationship are the cornerstones of medicine. We can confidently say cumulative medical radiation exposure appears to be positively associated with an increased risk of cancer, particularly haematological malignancies and solid brain tumours following CT. It is harder to confidently provide parents with the exact statistical chance of developing a malignancy given the number of confounding variables in the patient and our surrounding environment. Going forward, think about the ways you can reduce exposure to ionising medical radiation. Ask yourself:

“Is this test really needed? Will it change the patient’s diagnosis or management?”

“What imaging has the patient had previously? Is a repeat study necessary?”

“Can I organise an ultrasound or MRI for this patient, neither of which utilise ionising radiation?”

“When I do need to order a test with ionising radiation, what is the lowest possible dose that can be used to get the information I need?”

Ultimately, we should all be taking thorough histories and physical examinations. We should never order radiographic imaging blindly in order to “skip ahead” to a diagnosis. Instead, we should be using our clinical acumen to assess the need for investigations to help diagnose and manage patients. We must always aim to provide optimal care which is safe and practical.


Australian Radiation Protection and Nuclear Safety Agency (2021) Ionising Radiation and Health (online). Available at: [Accessed Nov 2021].

Berrington de Gonzalez et al. (2016) Relationship between paediatric CT scans and subsequent risk of leukaemia and brain tumours: assessment of the impact of underlying conditions, British Journal of Cancer, Volume 114 (4): 338-394

Berrington de Gonzalez et al (2004) Risk of cancer from diagnostic X Rays: estimates for the UK and 14 other countries, The Lancet, Vol 363 (9406), pp 345 – 351

Bibbo, G. (2018) Effective doses and standardised risk factors from paediatric diagnostic medical radiation exposures: Information for radiation risk communication, Journal of Medical Imaging and Radiation Oncology, Vol 62 (1): 43-50

Centers for Disease Control and Prevention (2015) The Electromagnetic Spectrum: Non-Ionizing Radiation (online). Available at [Accessed Nov 2021].

Centers for Disease Control and Prevention (2015) Measuring Radiation (online). Available at: [Accessed Nov 2021].

Hogan et al (2018) Radiation Exposure of Premature Infants Beyond the Perinatal Period, Hospital Paediatrics, Vol 11 pp 672-678

Mettler, F., Huda, W., Yoshizumi, T., Mahesh, M. (2008) Effective doses in radiology and diagnosis nuclear medicine: a catalog, Radiology, 248 (1):254-263, doi; 10.1148/radiol.2481071451 

Meulepas et al., (2019) Radiation Exposure from Paediatric CT scans and Subsequent Cancer Risk in the Netherlands, Journal of the National Cancer Institute, Vol 111 (3): 256 – 263

Oikarinen et al., (2019) Parents received and expected information about their child’s radiation exposure during radiographic examinations, Paediatric Radiology, Vol 49(2): 155-161

Pearce, M.S et al. (2012) Radiation exposure from CT scans in childhood and subsequent risk of leukaemia and brain tumours: a retrospective cohort study, The Lancet, Vol 380 (9840);499-504

Public Health England (2011) Ionising Radiation: dose comparisons (online). Available at:,the%20body%20to%20differing%20degrees. [Accessed Nov 2021].

The Royal College of Radiologists (2017) Paediatric Trauma Protocols. Available at:  [Accessed Nov 2021].

The United Nations Scientific Committee on the Effects of Atomic Radiation (2021). The Chernobyl accident (online). Available at: [Accessed Nov 2021}.

The United Nations Scientific Committee on the Effects of Atomic Radiation (2021) UNSCEAR Publications (online). Available at: [Accessed Nov 2021].

X Ray Risk (2018). Available at: [Accessed Nov 2021].

About the authors


High flow therapy – when and how?

Chest compressions in traumatic cardiac arrest

Searching for sepsis

The missing link? Children and transmission of SARS-CoV-2

Don’t Forget the Brain Busters – Round 2

An evidence summary of Paediatric COVID-19 literature


The fidget spinner craze – the good, the bad and the ugly

Parenteral Nutrition

Leave a Reply

Your email address will not be published.