Radiation protection and imaging stewardship — formative SAQs

SAQ 1 (10 marks)

A 6-year-old boy is being followed by the oncology service after treatment for a Wilms tumour. Over three years he has had multiple surveillance CT scans of the chest and abdomen. The registrar queries whether a routine surveillance CT is again due, and the parent has read that repeated CT scans "will give him cancer." You are asked to advise on the radiation-aware approach. [1] [8]

1. State the three pillars of radiological protection, and explain the examinable point about which of them does and does not apply to a patient investigation. (3) [8]
2. Quantify, in plain language suitable for the parent, the effective dose and the attributable lifetime cancer risk of a single paediatric CT, and name the cohort evidence that underpins the risk estimate. (4) [1] [2] [3]
3. Outline the stewardship approach to this child's surveillance imaging, including the dose-optimisation levers and the role of a non-ionising substitute. (3) [8] [10]

Model answer — SAQ 1

(1) The three pillars (3). The three pillars of radiological protection, set by the International Commission on Radiological Protection, are justification (any exposure must do more good than harm), optimisation (the ALARA principle — As Low As Reasonably Achievable), and dose limitation. The examinable point is that dose limits apply to occupationally exposed staff and to the public, and do not apply to the medical exposure of a patient whose investigation is justified — a justified patient investigation may exceed a staff dose limit, because the limit protects those who receive no benefit, whereas the patient receives the diagnostic benefit that justified the exposure. Justification and optimisation apply to every exposure; dose limitation does not apply to the patient. [8]

(2) The dose and the risk in plain language (4). A single paediatric head CT delivers an effective dose of the order of 1 to 2 mSv, comparable to several months of natural background radiation (around 2 to 3 mSv per year); an abdominal or chest CT is higher, of the order of 2 to 5 mSv and size-dependent. The attributable lifetime cancer risk is small and stochastic — probabilistic, with no safe threshold assumed (the linear-no-threshold model) — and is of the order of one excess cancer per several thousand to ten thousand scans. For the parent, I would frame this as a real but very small increase over the background lifetime cancer risk of roughly one in three, and emphasise that the risk of a justified scan is far smaller than the risk of missing a recurrence it is looking for. The evidence base is the New England Journal analysis of Brenner and Hall, the United Kingdom retrospective cohort of Pearce (leukaemia and brain tumours), the Dutch cohort of Meulepas, and the international EPI-CT consortium, all of which show a small but real dose-response and none of which forbid CT. [1] [2] [3]

(3) The stewardship approach (3). The first step is justification: is this surveillance scan needed now, and what will it change — the oncology surveillance protocol is followed, but each request is a deliberate decision, and the cumulative dose accumulated across the previous scans is part of the present decision. The second step is to consider a non-ionising substitute: MRI-based surveillance, where it answers the question, removes the ionising dose for that examination, and the shift towards MRI surveillance is the radiation-aware strategy. The third step, when CT is the justified modality, is dose optimisation: a size-specific paediatric protocol that scales the milliampere-seconds and kilovoltage peak to the child, a lower kilovoltage peak (typically 80 to 100 kVp), automatic tube current modulation, a limited scan length, a single phase rather than multiphase, and iterative reconstruction. The local protocol is benchmarked against the size-stratified paediatric diagnostic reference level (around the 75th percentile of the dose distribution, a benchmark not a limit). I would track the cumulative dose so that the next request is made against the lifetime total, and document the discussion with the parent. [8] [10]

SAQ 2 (10 marks)

A 4-year-old girl is referred for a pelvic radiograph for hip pain. The radiographer asks whether a gonadal shield should be applied. Separately, a colleague is about to request a multiphase abdominal CT for a child with suspected nephrolithiasis, and the department is reviewing its diagnostic reference levels. [8] [10]

1. State the current recommendation on routine gonadal shielding for this pelvic radiograph, and defend the reasoning against the traditional practice. (3) [12]
2. Defend or challenge the multiphase abdominal CT request on stewardship grounds, naming the dose quantities affected and the optimisation alternative. (4) [8] [10]
3. Define the diagnostic reference level, state how it is set and what it is not, and explain why it must be size-stratified in children. (3) [10]

Model answer — SAQ 2

(1) Gonadal shielding (3). Current ACR-AAPM and national guidance recommends discontinuing routine gonadal and fetal shielding for this pelvic radiograph. The reasons are technical: on modern equipment the dose outside the field of view is already negligible, a shield can slide into the beam and degrade the image, and a shield can trigger the automatic exposure control to raise the dose, so it may increase rather than reduce the dose. The shield also offers a false reassurance that distracts from the real optimisation levers of tight collimation and correct technique. Shielding of staff and of a pregnant carer who holds the child remains appropriate. The stewardship lesson is that protection is achieved by justification and optimisation, not by adding a shield to an exposure whose dose is already controlled. [12]

(2) The multiphase CT (4). I would challenge the multiphase request. The dose quantities affected are the computed tomography dose index (CTDIvol, in milligray), the dose-length product (DLP, in milligray-centimetres), and the effective dose in millisievert — and a multiphase protocol multiplies the dose by the number of phases, because each non-contrast, arterial, venous and delayed phase is a separate acquisition. For suspected nephrolithiasis, a single non-contrast phase is the diagnostic study that answers the question (it demonstrates the stone and any obstruction), so a multiphase protocol adds dose without changing management. The stewardship alternative is a single-phase, low-dose, non-contrast CT acquired with a size-specific paediatric protocol, a lower kilovoltage peak, automatic tube current modulation, a limited scan length, and iterative reconstruction. I would also ask whether a non-ionising modality — ultrasound, which can show hydronephrosis and, with twinkling artefact, a stone — answers the question before the CT is requested. The request names the clinical question, and only the phase that answers it is acquired. [8] [10]

(3) The diagnostic reference level (3). The diagnostic reference level (DRL) is a benchmarking tool for optimisation. For a defined examination on a defined patient group, it is set at around the seventy-fifth percentile — the third quartile — of the observed dose distribution across a sample of centres. It is not a dose limit and not a regulatory ceiling; it is an investigational trigger that flags a centre, a protocol or a technique whose typical dose is unusually high, so that the dose can be reviewed and reduced without sacrificing diagnostic quality. It must be size- or age-stratified in children, because a single adult-derived DRL applied to a child obscures the over-dosing the benchmark exists to detect — the size-specific dose estimate (SSDE) corrects the CTDIvol for the child's size, and the DRL is applied to that size-corrected distribution. [10]