[Theoretical bases for the critical reading of neurological articles on diagnostic tests]

Neurologists usually order diagnostic tests in clinical practice. A growing number and sophistication of the diagnostic tests in the last decades are evident. It is necessary for the clinician to know the operative functioning and accuracy of such tools. In addition, when reading a scientific paper about diagnostic test, the neurologist must pay attention to some methodological questions. These can be summarized as: a) blind and independent comparison between test result and gold standard, b) adequate gold standard, c) diagnostic test must be applied in appropriate spectrum of the disease, d) to appraise possible selection bias, and e) sufficient description of methodology, so it allow to replicate the study. Moreover, the evaluation should include data about test utility, and explain risk-benefit balance for the patients in whom it is used.