Using the guidelines

These guidelines are designed to assist the referrer in selecting the most appropriate investigation or procedure for a given diagnostic or imaging problem.


The guidelines are divided into the following sections:

  • Breast
  • Cancer
  • Cardiovascular
  • Chest
  • Gastrointestinal
  • Genitourinary
  • Head & Neck
  • Interventional 
  • Musculoskeletal
  • Neurological
  • Obstetrics & Gynaecology
  • Paediatrics
  • Trauma
  • Asymptomatic (Screening & Assessment)

Each section sets out the clinical scenario, including red flags, and lists relevant procedures with an indication of the associated radiation dose. For each procedure, there is a recommendation on its appropriateness (together with the grade). Finally, an explanatory comment is included when required to clarify the circumstances in which the procedure should be used.

Types of recommendations

The recommendations used are as follows:


Investigations most likely to contribute to the clinical diagnosis and management.

Specialised investigation.

Specialised investigations are frequently complex, time-consuming and/or resource-intensive, and will usually only be undertaken after discussion with the radiologist or according to locally agreed protocols

Indicated only in specific circumstances.

Non-routine investigations, usually only undertaken if a clinician provides cogent reasons or if the radiologist believes the examination represents an appropriate means of furthering the diagnosis and management of the patient. With certain clinical problems which may resolve with time, it may be correct to defer investigation.

Not indicated.

Investigations for which the proposed rationale is no longer appropriate.

The guidelines detail and provide links to key references and relevant clinical guidelines where appropriate.

Grades of recommendation

Clarification of evidence levels for diagnostic studies has been based on the levels of evidence for primary research adapted from the Oxford Centre for Evidence-based Medicine and The Journal of Bone and Joint Surgery (Table 1) 3,4. Recommendations have been graded according to evidence levels. Grading is adapted from the Oxford Centre of Evidence- based Medicine3. The highest level of evidence relevant to the clinical problem has been used to determine the grade of recommendation. In many instances a grade B or C reflects the supporting evidence base rather than the importance of these recommendations to the clinical problem addressed.

These grades are as follows:

[A] Any of the following

  • High-quality diagnostic studies in which a new test is independently and blindly compared with a reference standard in an appropriate spectrum of patients.
  • Systematic review and meta-analyses of such high- quality studies.

[B] Any of the following

  • Studies with a blind and independent comparison of the new test with the reference standard in a set of non- consecutive patients or confined to a narrow spectrum of patients.
  • Studies in which the reference standard was not applied to all patients.
  • Systematic reviews of such studies.

[C] Any of the following

  • Studies in which the reference standard was not objective.
  • Studies in which the comparison of the new test with the reference standard was not blind or independent.
  • Studies in which positive and negative test results were verified using different reference standards.
  • Expert opinion.

In some clinical situations there are conflicting data within a large body of excellent scientific reports. Thus, no firm recommendations are given, and the evidence is graded C. It should be noted that there are very few randomised, controlled trials that compare different radiological procedures. Assignment of evidence levels and grading of recommendations differs somewhat from those proposed by the Grading of Recommendations Assessment, Development and Evaluation GRADE Working Group 5 as supporting evidence is generally not from therapeutic studies but from diagnostic studies for which a Thornbury hierarchy may be more relevant.

Table 1.3,4
Level 1
Therapy/ Prevention, Aetiology/ HarmSystematic review (SR) of randomised control trials (RCT). Individual RCT. All or none studies (eg, when all patients died before the intervention became available)
PrognosisSR of inception cohort studies; Clinical decision rules (CDR) validated in different populations. Individual inception cohort study with >80% follow-up; CDR validated in a single population. All or none case-series.
DiagnosisSR of Level 1 diagnostic studies; CDR with studies from different clinical centres. Validating cohort study with good reference standards; or CDR tested within one clinical centre. Absolute SpPins and SnNouts
Differential Diagnosis/ Symptom prevalence studySR of prospective cohort studies. Prospective cohort study with good follow-up. All or none case-series.
Economic and decision analyseSR of Level 1 economic studies. Analysis based on clinically sensible costs or alternatives; systematic review(s) of the evidence; and including multi-way sensitivity analyses. Absolute better-value or worse-value analyses.

Level 2
Therapy/ Prevention, Aetiology/ HarmSR of cohort studies. Individual cohort study. ‘Outcomes’ research; ecological studies.
PrognosisSR of either retrospective cohort studies or untreated control groups in RCTs. Retrospective cohort study or follow-up of untreated control patients in an RCT; Derivation of CDR or validated on split-sample only. ‘Outcomes’ Research.
DiagnosisSR of Level >2 diagnostic studies. Exploratory cohort study with good reference standards; CDR after derivation, or validated only on split-sample or databases.
Differential Diagnosis/ Symptom prevalence studySR of 2b and better studies. Retrospective cohort study or poor follow-up. Ecological studies.
Economic and decision analyseSR of Level >2 economic studies. Analysis based on clinically sensible costs or alternatives; limited review(s) of the evidence, or single studies; and including multi-way sensitivity analyses. Audit or outcomes research.

Level 3
Therapy/ Prevention, Aetiology/ HarmSR (with homogeneity*) of case-control studies. Individual case-control study.
DiagnosisSR of 3b and better studies. Non-consecutive study; or without consistently applied reference standards.
Differential Diagnosis/ Symptom prevalence studySR of 3b and better studies. Non-consecutive cohort study or very limited population.
Economic and decision analyseSR of 3b and better studies. Analysis based on limited alternatives or costs, poor quality estimates of data, but including sensitivity analyses incorporating clinically sensible variations.

Level 4
Therapy/ Prevention, Aetiology/ HarmCase-series (and poor-quality cohort and case-control studies).
PrognosisCase-series (and poor-quality prognostic cohort studies).
DiagnosisCase-control study, poor or non-independent reference standard.
Differential Diagnosis/ Symptom prevalence studyCase-series or superseded reference standards.
Economic and decision analyseAnalysis with no sensitivity analysis.

Level 5
Therapy/ Prevention, Aetiology/ HarmExpert opinion without explicit critical appraisal, or based on physiology, bench research or ‘first principles’.
PrognosisExpert opinion without explicit critical appraisal, or based on physiology, bench research or ‘first principles’.
DiagnosisExpert opinion without explicit critical appraisal, or based on physiology, bench research or ‘first principles’.
Differential Diagnosis/ Symptom prevalence studyExpert opinion without explicit critical appraisal, or based on physiology, bench research or ‘first principles’.
Economic and decision analyseExpert opinion without explicit critical appraisal, or based on economic theory or ‘first principles’

Grades of Recommendation (Adapted from Oxford Centre for Evidence-based Medicine – Levels of Evidence (March 2009)3.

AConsistent level 1 studies
B Consistent level 2 or 3 studies or extrapolations from level 1 studies
C Level 4 or 5 studies or extrapolations from level 2 or 3 studies or troublingly inconsistent or inconclusive studies of any level

Radiation Dosage

In these referral guidelines, the doses are grouped into broad bands to support referrers in their knowledge of the amount of radiation dose involved for various investigations (Table 3)

Table 3. Band classification of the typical doses of ionising radiation from common imaging procedures14, 17

SymbolTypical effective dose (mSv)* Examples Lifetime additional risk of cancer induction per exam for adults*
None 0Ultrasound, Magnetic Resonance Imaging 0
Radiationicon< 1 Chest, extremity, lumbar spine X-ray NM GFR (Tc-99m) <1:20,000
RadiationiconRadiationicon1–5 CT head, NM bone scan (Tc-99m), coronary angiography 1:20,000 to 1: 4,000
RadiationiconRadiationiconRadiationicon5.1–10 CT pulmonary angiogram, NM cardiac SPECT(Tc-99m)) . 1:4,000 to 1: 2,000
RadiationiconRadiationiconRadiationiconRadiationicon> 10 CT chest, abdomen & pelvis, whole body PET-CT (F-18-FDG) >1:2,000

*Cancer risk indicated in this table is averaged for adults. The risks from radiation vary considerably with age and sex, with higher risks in children and females 17. As risks for children are higher, the examinations indicated may need to be moved to higher risk band, for example a CT head for a child may move into a RadiationiconRadiationiconRadiationiconrisk band.

General information

Imaging Techniques

Computed Tomography (CT)

CT images are produced by an X-ray tube rotating rapidly around a patient, producing a three-dimensional dataset that may be manipulated to examine different structures, in different planes with high resolution (typically under 1 mm). CT protocols are tailored to optimise the administration of oral and/or intravenous contrast medium to answer the clinical question(s) posed.

The major drawback of CT is the relatively high radiation dose. The amount of radiation delivered may be reduced by a number of technologies including automatic exposure control, low dose protocols, iterative reconstruction and dual energy imaging. The number of phases of imaging should also be minimised to reduce exposure.

The use of CT in screening should be directed at those groups at high risk where there is supporting evidence- the use of targeted CT for selective screening in national programmes has been shown to be beneficial.

At present, there is no evidence to support general population health screening with CT and the use of whole-body CT for ‘screening’ of self-presenting, asymptomatic subjects is not justified.

Interventional Radiology (IR)

Image-guided procedures are used to diagnose and/ or treat both non-vascular and vascular conditions. These include:

  • Biopsy
  • Drainage of fluid collections/ obstructed ducts
  • Re-vascularisation of blocked vessels
  • Occlusion of bleeding vessels
  • Treatment of focal lesions, e.g. tumour

Almost all specialties within an acute hospital use interventional radiology (IR) services. The provision of therapies by IR must be considered in terms of clinical appropriateness, local expertise and patient choice. Close collaboration with clinical colleagues and a robust system for patient information and consent are essential. While the development of management algorithms is encouraged, clinicians are urged to discuss cases with the radiology department. Radiologists, clinicians and patients should agree on treatment, expected risks and outcomes. The boundary between diagnostic and interventional radiology is blurred and differs between hospitals/units. Interventional radiologists will often oversee the imaging investigation of conditions they manage. In some departments, biopsy and drainage are performed by IR, while in other departments diagnostic radiologists will assume this role.

Not every hospital will be able to offer every treatment. Where IR services are not available in departments, there should be formal arrangements with other hospitals that allow patients access to services.

Magnetic Resonance Imaging (MRI)

MRI involves placing patients in a strong magnetic field and applying a series of radiofrequency pulses. The signal returned from the tissues is reconstructed to generate images. The great advantage of MRI is the absence of any ionising radiation.

Consequently, MRI should be considered in preference to CT when both investigations could yield similar information. The major uses for MRI are in evaluation of neurological syndromes and musculoskeletal disorders, together with oncological, hepatobiliary and gynaecological disease.

MRI may be contraindicated in patients with metallic foreign bodies in situ; for example, within the orbits. All implanted devices should be declared to the department in advance. All patients with implanted devices should be discussed with the MRI department if there is any concern of compatibility or if implantation is very recent (<6 weeks). In addition to safety issues, image quality is often degraded close to implants.

The safety of MRI during the first trimester of pregnancy is uncertain. However, it may well be safer than some of the alternative options. Always notify the imaging department if pregnancy is suspected or confirmed.

Nuclear Medicine

NM provides a functional approach to imaging of pathology by combining a physiological agent attached to an isotope– most commonly Technetium-99m (Tc-99m). Radiation dose is similar to and often less than for comparable anatomical studies (Table 3).

Hybrid imaging combines anatomy derived from CT images with functional imaging from the radioisotope and is imaged by dedicated single photon emission computed tomography (SPECT-CT) scanners.

Defining the precise clinical problem and question to be answered on the request form will assist radiologists and nuclear medicine physicians in the selection of the most appropriate procedure.


Hybrid imaging with positron emission tomography (PET) co-registered with CT (PET-CT) combines pathophysiological information, with anatomical localisation. PET utilizes radiopharmaceuticals which decay with positron emission and annihilation producing simultaneous gamma rays at 180 degrees to each other thus improving signal-to-noise ratio and localisation. Low-dose CT data, obtained during the same examination, are co-registered with positron activity. Consequently, the effective radiation dose from PET-CT is comparatively high, as noted in Section 3.

The main application of PET-CT is in the staging, follow-up and restaging of patients with cancer using the glucose analogue 2-[18]fluoro-2-deoxy-D-glucose (FDG). FDG PET- CT identifies the increased glucose metabolism seen in malignant cells compared with non-malignant cells.

A number of other radiotracers are available to evaluate other physiological processes, but this varies significantly between centres.

With regard to PET-CT within the guidelines, specific radiopharmaceuticals have not generally been defined, as the isotope used and technology (including SPECT- CT) will change with time. Justification and optimisation decisions are dependent therefore on local protocols and availability.


The absence of ionising radiation and its increasing portability make US the first-choice investigation for a wide range of clinical conditions. US remains highly operator and patient-dependent and requires skills which take time to acquire. It must be undertaken by trained, experienced operators but there are limitations to images obtained and also the conclusions which can be drawn in every case.

The central role of US in image-guided biopsy and other interventional techniques has contributed to its growth. The use of image guidance for many procedures is mandatory for safe clinical practice; for example, central venous l line placement, or in pleural fluid aspiration.


Increasingly, US is utilized outside of the imaging department by a wide range of practitioners. Wherever possible, the images should be recorded together with a report of the interpretation of the images, as with any other investigation.

Radiography and Fluoroscopy

Radiographs remain the first line imaging investigation for many conditions, particularly for chest and musculoskeletal problems, especially suspected fractures. The inherent low dose and relatively low cost mean that for many clinical questions, radiography is the ideal investigation to provide a definitive answer. Radiographic images have the advantage of showing easily recognisable anatomy enabling initial interpretation by appropriately trained clinicians. As for all imaging investigations, a report is required and will give added value.

Using X-ray image intensification, fluoroscopy enables the production of real-time video images which are used in a number of clinical applications, including diagnostic contrast studies, guidance for intervention and surgical procedures. Use by appropriately trained operators will avoid high radiation dose which can result from prolonged fluoroscopic screening.

Contrast Agents

Intravascular iodinated contrast agents have been used safely in contrast studies, CT and interventional radiology for many years with a very low incidence of serious reactions (0.004%) 19

In patients with moderate or severe anaphylactic reactions, formal skin testing may be advisable as this is commonly brand-specific and may have significant implications for patients. Contrast-induced nephropathy is much less common than previously thought and should be considered rare with intravenous infusions 19.

Gadolinium-based intravenous contrast agents used in MRI are generally well tolerated and have a good safety record in general. However, caution should be exercised in patients with impaired renal function as these patients are at risk of nephrogenic sclerosing fibrosis. The incidence varies between brands of contrast agents and appears to be related to the molecular structure 19.

Gadolinium deposition within the brain and other organs has been reported although the clinical significance is unclear as there are no associated clinical syndromes described.

Ultrasound contrast media comprises microbubbles of gas surrounded by a ‘shell’ that differs between various brands. These are well-tolerated with no significant side effects reported 19.

Oral or enteral contrast agents such as barium sulphate or water-soluble iodinated agents are rarely associated with adverse reactions. Nausea, vomiting and constipation may be reported, especially in dehydrated patients 19.

Considerations for specific patient cohorts


Pregnancy and Protection of the Foetus

Irradiation of a foetus should be avoided whenever possible. This includes situations in which the woman herself does not suspect pregnancy. The prime responsibility for identifying such patients lies with the referring clinician, but radiology staff must also check the pregnancy status of patients when they attend for examination (not least because some time may have elapsed since the clinician completed the request card).

Women of reproductive age presenting for an examination in which the primary beam irradiates the pelvic area (essentially, any ionising irradiation between the diaphragm and the knees), directly or by scatter, or for a procedure involving radioactive isotopes, should be asked whether they are or may be pregnant. If a patient cannot exclude the possibility of pregnancy, she should be asked if her period is overdue.

If the patient can exclude the possibility of pregnancy, the examination can proceed. The examination can also go ahead where pregnancy cannot be excluded, but the period is not overdue and the investigation carries a relatively low dose to the uterus. However, where pregnancy cannot be excluded, the period is not overdue but the examination carries a high dose, the procedure outlined below should be followed. ‘High dose’ in this context is defined by the Health Protection Agency (HPA) as any examination falling in the highest dose group resulting in foetal doses of more than about 10 mGy 18. Typically, the only investigations in this category will be pelvic or abdominal CT, complex fluoroscopy and a few nuclear medicine procedures.

If the patient is definitely pregnant, or if pregnancy cannot be excluded and the period is overdue, or the patient is not conscious, the justification for the proposed examination should be reviewed by the radiologist and the referring clinician, with a decision taken on whether to defer the investigation until after delivery or until the next menstrual period has occurred. However, a procedure of clinical benefit to the mother may also be of indirect benefit to her unborn child, and a delay in an essential procedure may increase the risk to the fetus as well as to the mother. This consideration is especially relevant in the emergency situation. In all cases, if the radiologist and referring clinician agree that irradiation of the pregnant or possibly pregnant uterus is clinically justified or is not clinically justified, this decision should be recorded and the patient must be kept fully informed. If it is decided that the irradiation is justified, the radiologist must ensure that exposure is limited to the minimum required to acquire the necessary information.

If it becomes obvious that a fetus has been inadvertently exposed, despite the above measures, the small risk to the fetus, even at higher doses, is unlikely to justify the greater risks of invasive fetal diagnostic procedures (for example, amniocentesis) or those of termination of the pregnancy. When such inadvertent exposure has occurred, a medical physicist should make an individual risk assessment and the results should be discussed with the patient.


Children are more sensitive than adults to the harmful effects of ionising radiation. Exposure to ionising radiation and invasive diagnostic or interventional procedures should be avoided whenever possible

Plain radiography and fluoroscopic contrast examinations have an important role in paediatric diagnosis but should be used only when clinically necessary, and exposure factors should be appropriate to the size of the child. In fluoroscopic examinations, screening times should be kept as low as possible.

Ultrasound is often the first modality of choice for imaging children as there is no radiation involved. Ultrasound examination is painless, non-invasive and, with patience, useful images can be obtained even with a restless child. However, ultrasound cannot penetrate bone and air, thus some anatomical locations are not accessible to ultrasound examination.

Magnetic resonance imaging (MRI) allows non-invasive imaging with no exposure to ionising radiation, providing a high level of intrinsic contrast and the capability of multiplanar imaging. MRI does not offer the same capability for real-time imaging that ultrasound can provide; imaging times are relatively long, and as the images are easily degraded by movement artefact, younger children may require sedation or general anaesthesia. However, MRI is not constrained by the same limitations as ultrasound to anatomically accessible areas; the presence of bone and air does not interfere with MR images.

CT involves exposure to relatively high doses of ionising radiation and alternative modalities should always be considered, but CT is often the modality of choice in major trauma and other emergency situations where rapid cross- sectional imaging is necessary. It is important to ‘right size’ the exposure factors, to scan only the essential anatomical areas and to avoid multiple exposures.

Radionuclide imaging can provide unique functional information but is generally indicated in children only in the special circumstances listed in these guidelines.