It is difficult to know whether I am a typical radiologist, a typical doctor, or even a typical man, but when it comes to screening, I am very much of the ostrich variety. I eschew the opportunity to discover my prostate specific antigen and cholesterol levels, and the very idea of a fecal occult blood test I find faintly obscene.

But it is impossible to ignore the issue of screening because it threatens to engulf the imaging specialties. Imaging is now at the forefront of the quest to root out disease: mammography, aortic ultrasound, nuchal thickness measurement, calcium scoring, lung cancer detection, imaging for microscopic hematuria, whole-body imaging, carotid artery imaging, and bone densitometry. No risk of unemployment for the radiologist, then!

The issues of screening, however, have never been more complex. Screening works when the end point is clear: find occult disease, treat disease, improve outcome. But the defining principles are changing.

Increasingly, the outcomes of screening are influenced by choice. Parents may choose to avoid screening for Down syndrome, or they may simply wish to know about the level of risk so they can prepare for the possible birth of a child affected by Down. Alternatively, they may be totally unprepared to accept any risk of an affected child and choose invasive testing irrespective of the screening result. But if screening is about choice and information, then the data about efficacy, quality-adjusted life years, and cost-effectiveness become skewed.

Some individuals avoid all forms of screening and guarantee immortality by relying instead on the daily glass of red wine, copious amounts of garlic, olive oil, and whole grain cereal, and the pain and exhaustion provided by a session in the gym. Others will be reassured only by availing themselves of all services designed to detect every disease.

“Intima medial thickness measurements? Yes, please! Whole-body scanning? That's for me! But surely PET/CT would be better at detecting occult neoplasia?”

How do doctors influence this process? A number of articles have suggested that we may influence the uptake of screening unfairly. Even established programs with an extensive evidence base, such as breast screening, can be challenged on the basis of patients' lack of information. That is not to say that information is not provided, simply that the information may appear to push them toward choosing to be screened rather than choosing not to be screened. Information leaflets may ask “Why do I need breast screening?” rather than “Do I need breast screening?”

The advantages of early diagnosis are offset by unnecessary or inappropriate treatment. If we are to provide choice, then we have to provide complete information. With choice, however, come challenges to the evidence base. If I were to be screened for the level of my cholesterol and if screening were to alter the outcome, I would need to alter my behavior if the level is raised. I enjoy my butter, bacon, cheese, and red meat far too much to allow that to happen.