Purpose: In general, articles on diagnostic tests have a very poor methodological quality. If we translate their conclusions to daily practice without a proper analysis it is easy to see a typical trend: physicians use new (and expensive) tests without increasing diagnostic capacity; they only increase the health budget. Evidence-based medicine (EBM) consists of using the best evidence in decision-making. It deals with selected and relevant information, supported by data obtained through the most rigorous scientific method: epidemiology and statistics. Critical evaluation of papers related to diagnostic tests is our aim. We provide with basic skills for evaluation and analysis of papers by means of simple topics on statistics, design of clinical assays and clinical epidemiology.
Methods: Review of the medical literature.
Results: To evaluate papers on diagnostic tests and to use a test correctly, it is necessary to know its diagnostic capacity, the level of certainty to start treatment, the probability of having the disease before using the test and the test capacity to change that probability.
Conclusions: The maximum performance of a test is reached when it is used in a maximum uncertainty situation (disease probability 50%). Its highest usefulness is obtained when pre-test probability is maximal, after a careful review of the patient's history and a complete patient examination.