Review nd complemented by popular tools such as the Graphic the need to identify poor ractices in how of biased research ptimal anaesthe ming that" ch findingfs]are false e of ger ration of rese quarter century.The current ch wasted:"th within afa ork of its his of women e ma portan sed Promotion prophylact add the sand limita han infarcti of these drugs than in the Vietnam waMillions of 9 ealthy wo were pre between medical evidence,theory.a trials refuted thes ssed as a cohere increased the inid tructure for optimising e practice of medicin EBM. inception oped ogy of EBm cting the first EBM epis gical principle:the is der ermined b edibl tecicsofheaihinteneaiom quantitative results of clinic This work. sed on front-line of medicine should be on the part of the core requirements for most view that the pursu of truth is best The initial hierarchies of ating sign vidence is.howe ece ary but not rega ding the sup eriority res of i ortance to the the tr and although early work ful decision making requires consideration of patients ofter values and preferences on surroga ma EBM's initial hierarchy of evidence 16.2017http他，dom10.1016/s0140-661631592.6 Review 2 www.thelancet.com Published online February 16, 2017 http://dx.doi.org/10.1016/S0140-6736(16)31592-6 understanding the results of clinical studies, and determining how best to apply the results to their everyday practice.9 Subsequently, detailed guidance published in journal articles and associated textbooks,10,11 complemented by popular tools such as the Graphic Appraisal Tool for Epidemiology,12 resulted in EBM becoming increasingly integrated into medical curricula worldwide.13 Additionally, ratings of important developments in medicine have placed EBM on par with antibiotics and anaesthesia for the practice of medicine.14 Here, we briefl y review the philosophical underpinnings of EBM and, in more detail, its progress during the past quarter century. The current discussion goes beyond previous reviews,15 placing the development of EBM within a framework of its historical and philosophical underpinnings; highlighting the role of EBM in the development of standards for clinical research and measuring practice; clearly documenting the important changes in EBM that occurred over more than two decades; addressing the critiques and limitations of EBM; and predicting the development of EBM in the next 25 years. EBM and the theory of knowledge On the surface, EBM proposes a specifi c association between medical evidence, theory, and practice. EBM does not, however, off er a new scientifi c theory of medical knowledge,16,17 but instead has progressed as a coherent heuristic structure for optimising the practice of medicine, which explicitly and conscientiously18 attends to the nature of medical evidence. Central to the epistemology of EBM is that what is justifi able or reasonable to believe depends on the trustworthiness of the evidence, and the extent to which we believe that evidence is determined by credible processes.17 Although EBM acknowledges a role for all empirical observations, it contends that controlled clinical observations provide more trustworthy evidence than do uncontrolled observations, biological experiments, or individual clinician’s experiences. The basis for the fi rst EBM epistemological principle is that not all evidence is created equal, and that the practice of medicine should be based on the best available evidence. The second principle endorses the philosophical view that the pursuit of truth is best accomplished by evaluating the totality of the evidence, and not selecting evidence that favours a particular claim.16 Evidence is, however, necessary but not suffi cient for eff ective decision making, which has to address the consequences of importance to the decision maker within the given environment and context.17 Thus, the third epistemological principle of EBM is that clinical decision making requires consideration of patients’ values and preferences. EBM’s initial hierarchy of evidence EBM originally focused on documenting biases in research applied to clinical practice, understanding the results of clinical studies, and considering situations (related to patient characteristics, family, and social and economic environment) in which these results can and cannot be usefully applied. In doing so, EBM addressed the need to identify poor research practices in how research is conceived, conducted, published, and used. Several investigators have provided examples of biased research leading to suboptimal medical practice, lamenting the “scandal of poor medical research”19 and claiming that “most research fi nding[s] are false”.20 Estimates suggest that 50% of research eff ort is wasted at each stage of generation and reporting of research, resulting in more than 85% of total research wasted; 21 the human toll of spurious research fi ndings has equally been enormous.22 For example, thousands of women underwent gruelling and sometimes fatal bone marrow transplantation for treatment of breast cancer based on biased research.23 Promotion of prophylactic antiarrhythmic therapy in patients with myocardial infarction, based on physiological reasoning that suppression of arrhythmias would reduce mortality, proved disastrous—more Americans died from the use of these drugs than in the Vietnam war.24 Millions of healthy women were prescribed hormone replacement therapy on the basis of hypothesised reduction in cardiovascular risk; randomised trials refuted these benefi ts and demonstrated that hormone replacement therapy increased the incidence of breast cancer.25 In response, EBM, from its inception, developed schemas for the assessment of the quality of evidence, refl ecting the fi rst EBM epistemological principle: the higher the quality of evidence, the closer to the truth are estimates of diagnostic test properties, prognosis, and the eff ects of health interventions. Further, EBM writings acknowledged the challenges of understanding the quantitative results of clinical research, and of applying these results to patients who do not necessarily fi t the eligibility criteria of the available studies. This work, focused on educating front-line clinicians, was so quickly acknowledged that, within a decade of their introduction, EBM principles became part of the core requirements for most undergraduate and postgraduate medical education worldwide. The initial hierarchies of evidence that EBM proposed focused on the design of clinical studies and were relatively simple (fi gure 1A). For therapy, the hierarchy provided no equivocation regarding the superiority of randomised controlled trials (RCTs) over observational studies for determining the trustworthiness of evidence related to treatment eff ects—although early work fully acknowledged the limitations of small sample size and the questionable application of clinical fi ndings, often based on surrogate markers, to patients who diff ered from those included in the primary studies. Almost immediately, observers objected, noting that RCTs can also be biased, and hence should not automatically be equated with high-quality evidence.28 As