[Can therapeutic trials serve as a model for the evaluation of diagnostic methods?]

Many health care professionals think that diagnostic technologies must be assessed as treatments are and that controlled diagnostic trials must be organized for evaluation of diagnostic tests. There are many differences between these two types of trials relating with: unfeasibility of blind judgments, biases due to requesting of diagnostic tests in addition to technologies to be compared, number of patients required for a sufficiently powerful statistical analysis, diverging interest of the various clinical teams involved into the assessment. A different approach using simulation methods and decision trees appears more appropriate in the evaluation of diagnostic technics.