How well does "evidence-based" medicine help neurologists care for individual patients?

The advocates of evidence-based medicine have established a clear, unambiguous requirement for what they consider credible evidence: the randomized, controlled trial and systematic reviews of several or many randomized, controlled trials. However, evidence from therapeutic trials cannot always be applied to the care of individual patients. These studies yield information only on the likely benefit of a particular treatment strategy among a large group of often heterogeneous patients with a given condition. Some trials may show statistically significant positive results for a certain therapy, but these results may be of little or no practical benefit for most patients with the condition studied. Other studies may show that a certain treatment is beneficial to most patients, although it may have serious risks and cause harm in others with the same condition. To be therapeutically useful, a trial must study specific, well-defined, and quantifiable conditions with important and readily recognized and defined endpoints. More importantly, physicians should spend more time finding out what is wrong with each patient and getting to know his or her circumstances, family situations, psychosocial and economic stresses, thoughts, fears, biases, and wishes.