How good is the evidence base for test selection in clinical guidelines?

• The concepts of guidelines and test selection are reviewed.
• The ideal evidence base for test selection is explored.
• Challenges and sources of bias in evidence for test selection are discussed.
• Challenges in implementing and adhering to guidance are discussed.

Clinical guidelines are ubiquitous, manifold and form an integral component of evidence-based clinical practice. Guidelines on test selection are often considered a useful adjunct to aid clinical decision-making, as test selection is a complex process that is influenced by many patient, clinician and laboratory factors. However, it is important to carefully evaluate several aspects of these guidelines, which include the context of the test in the guideline, the quality of the studies underpinning recommendations, the extent of the evaluation of effectiveness (or performance) of the specific test and in the clinical pathway, its applicability and ease of implementation. A robust evaluation of a diagnostic test should incorporate several stages including evaluation in healthy, symptomatic but unaffected and affected populations, and importantly a measurement of impact on patient outcomes. Few diagnostic studies meet these criteria, and therefore crucial aspects of test evaluation are overlooked prior to incorporation into clinical guidelines. Whilst efforts are made to standardise reporting of studies, strength of evidence and quality of guidelines, further work is required to improve the quality of the diagnostic studies that formulate these guidelines. It is important that clinicians using guidelines for test selection appreciate the limitations of the diagnostic test, and the guidelines themselves.