With the development of multiple imaging techniques such as ultrasound, computed tomography, magnetic resonance imaging, and photon emission tomography in addition to conventional radiography, contrast radiography and vascular radiology it is important that clinicians have guidance on the appropriateness and merits of these techniques. A useful test is one whose result—positive or negative—will alter patient management; many requests for imaging do not. Each unnecessary test increases costs, may add to patient irradiation, blocks the service, lowers standards, and harms morale. Other than for medicolegal reasons the chief causes of wasteful use of tests are:

1.Imaging when results are unlikely to affect patient management because the anticipated ‘positive’ finding is usually irrelevant (e.g. degenerative spinal disease) or because a positive finding is so unlikely.

2.Examining too often, i.e. before the disease could have progressed or resolved or before the results will influence treatment.

3.Imaging when it has already been done, e.g. at another hospital, in outpatient or accident and emergency clinics.

4.Failing to explain the purpose of the examination, so that the wrong films are taken or an essential view is omitted.

5.Doing the wrong study; imaging techniques change all the time.

A guideline is ‘not a rigid constraint on clinical practice, but a concept of good practice against which the needs of the individual patient can be considered’. So while there have to be good reasons for ignoring guidelines they should not be regarded as absolute rules. No set of guidelines will command universal support and any problems should be discussed with your own radiologists. The guidelines discussed here reflect the practice in the United Kingdom. They have been adapted for local needs and surgical practice from the second edition of the Royal College of Radiologists' (UK) Guidelines. Copies of the Royal College of Radiologists Guidelines can be obtained from the Royal College of Radiologists, 38 Portland Place, London, W1N 3DG, UK. They may vary from one hospital to another and from one country to another. They may need to be adapted to suit local circumstances.

Although X-rays are taken for granted there is no known safe dose. Some ‘spontaneous’ genetic mutations and some malignancies are attributable to background radiation and a lumbar spine radiograph gives a radiation dose equivalent to a year's background radiation. Radiation must be taken seriously.

CT examinations are high dose examinations. They account for 2 per cent of all examinations using X-rays in the United Kingdom but contribute 20 per cent of the population's radiation dose from diagnostic X-rays.

In view of the potential for high patient doses of X-rays, CT examination should only be carried out after there has been proper clinical justification for the examination of each individual patient by an experienced radiologist. Examinations of children require a higher level of justification, since they are at greater risk from radiation than are adults. When clinically appropriate, the alternative use of safer non-ionizing techniques (such as ultrasound and MRI) or of low dose X-ray techniques should be considered.

Could the patient be pregnant?

Irradiation of any area between the diaphragm and the knees should be avoided in pregnancy unless there are over-riding clinical considerations. If the patient is, or might be, pregnant the department of radiology must be informed.

In general these guidelines only deal with areas of difficulty or controversy. Straightforward indications for examinations are not discussed, nor are the indications for examinations where the requests are routinely evaluated by a radiologist, e.g. CT, MRI, or arteriograms.

The sections are laid out by systems with further sections for accident and emergency, trauma, paediatrics, ultrasound, mammography, and nuclear medicine. There are three columns. The first gives the clinical situation which may be the basis for requesting an examination. The next column is the guideline on whether or not the investigation is appropriate, and the third provides explanatory comments. The guidelines used are:

1Indicated. This guideline is used to indicate the most appropriate investigation for that clinical situation, and may differ from the examination requested.

2Six-week rule. These are situations in which experience shows that the clinical problem usually resolves with time and we therefore suggest deferring the study for 6 or 8 weeks and only performing it if the symptoms are still a problem. Acute back or neck pain are common examples.

3Not indicated routinely. This emphasizes that while no guidelines are absolute the request will only be carried out if a clinician gives strong arguments for it. An example would be a patient with backache in whom there were findings to suggest disease other than degenerative disease.

4Not indicated. Examinations in this category are those where the supposed rationale for the test is untenable. A cervical spine radiograph may be requested in patients with suspected vertebrobasilar insufficiency ‘to show degenerative changes’. In fact all patients in this age group have degenerative changes and it is impossible to know whether or not those changes affect the vertebral arteries.

5Specialist clinicians only. These are studies that will only be performed for consultants who have relevant clinical expertise to evaluate the clinical findings and act on the imaging results.

6Limited examination. Here the study is confined, as explained above, usually to a single view intended to show any major abnormality, e.g. lateral skull for epilepsy in the absence of localizing signs.

Mammography is indicated for the diagnosis, preoperative assessment, and management of breast cancer.

Where there is clinical suspicion of breast cancer, e.g. a suspicious lump, residual lump following aspiration of a cyst, or a single duct discharge, then referral to a breast clinic/breast surgeon is advisable PRIOR TO ANY RADIOLOGICAL INVESTIGATION.

1.Breast pain and/or tenderness in the absence of other clinical signs;

2.Generalized breast lumpiness;

3.Long-standing nipple inversion;

4.Symptoms related to the contraceptive pill.

1.As a routine prior to or during hormone replacement therapy;

2.In the management of cyclical mastalgia without clinical signs;

3.As the firstline investigation in women under 35 unless there is a STRONG clinical suspicion of breast cancer. The sensitivity of mammography in detecting malignancy is reduced in the young glandular breast and cancer is uncommon in this age group.

The radiologist may decide that ultrasound is a more appropriate method of investigation to solve the clinical problem. Ultrasound is not suitable for screening at any age.

Remember that although mammography is the best method of detecting early breast cancer, it is not 100 per cent sensitive. A negative mammogram does not exclude breast cancer. All imaging results must be considered in the context of the clinical findings.

Routine mammography may be justified at any age in women who are at high risk of developing premenopausal breast cancer. This includes those with histological risk factors from previous surgery and first-degree relatives of those who have had premenopausal breast cancer. Mammography should be supervised by a breast clinic or specialist.

Local policies and availability vary and general guidelines are not appropriate. Many departments and screening units offer a service only through prior referral to a breast clinic or specialist.

Women with a mother or sister who developed breast cancer before menopause, or in whom two or more close relatives (mother, sister, aunt, grandmother) developed breast cancer, are at higher risk. Appropriate counselling and a personal plan for mammography may be required. If the family history is of postmenopausal breast cancer, referral to the clinic or specialist may be delayed until the age of 35. Where the family history is of premenopausal cancer, referral at an age 10 years younger than the first-degree relative who developed breast cancer may be advisable.

A single first-degree relative who develops postmenopausal breast cancer confers only a slight increased risk to others and routine mammography is not usually indicated in women under the age of 50 .