This chapter discusses the meaning of evidence-based medicine and where it relates to randomised controlled trials, but also where it does not. The need for good quality evidence is stressed through a discussion of high failure rates in drug development and arguments against access to unlicensed (and largely untested) treatments are set out (despite the good intentions of those who advocate such access to treatments).Good quality, reliable evidence does not always have to come from clinical trials. Other forms of evidence are discussed. Meta-analyses of individual trials may help to resolve the problem that, in rare diseases, it may be very difficult or impossible to do adequately powered clinical trials - but that does not imply those trials have no value at all.The importance of patients' choices is stressed but the difficulties of making choices and the general poor understanding of risk can make patients and caregivers, as well as healthcare professionals, very vulnerable to making poor decisions. All stakeholders need to be adequately guided through the evidence to make proper informed decisions.
Keywords: Bias; Bradford Hill; Evidence; Meta-analysis; Patient preference; Precision.