For purposes of this discussion, the major criticisms that have been raised are presented here as five main themes.
Explore This IssueOctober 2008
Perhaps the most fundamental of these criticisms addresses a veiled empiricism that may philosophically underpin the concept of EBM. A basic tenet of EBM is its elevation of experimental evidence over pathophysiological or other forms of knowledge. In order to achieve this, it is assumed that the scientific observations on which experimental evidence is based can be made independent of the theories and biases of the observer. However, the impossibility of bias-free observation has been recognized for centuries. In the late 19th century, French mathematician and physicist Pierre Duhem argued that science enables a large number of empiric facts to be related to one another through use of a relatively small number of principles, but the observations themselves cannot be carried out by a naive observer.8 In other words, all observations are influenced by the perceptions and beliefs of the observer, and therefore bias can never be eliminated. Using this reasoning, critics suggest that EBM may, in fact, unknowingly create an environment of over-reliance in the power of empiricism to determine what constitutes reliable knowledge.
Definition of Research Quality
The next critical theme addresses the way in which research is defined and graded. Within EBM, the definition of evidence is relatively narrow and tends to devalue some forms of information that may be of importance within the health care arena.9 Research methodology is used as a basis for EBM grades. Those research methodologies thought to be less vulnerable to bias are graded as higher, and those thought to be more vulnerable are graded as lower.10 However, to date there is no evidence to suggest that higher EBM grades of evidence are any more reliable than lower grades. In 2000, Benson and Hartz compared observational reports addressing 19 diverse treatments with randomized, controlled trials for the same treatments. They found little evidence that estimates of treatment effects in observational studies reported after 1984 are either consistently larger than or qualitatively different from those obtained in randomized, controlled trials.11 Further, the meta-analyses and randomized controlled trials that constitute the highest grade of research may not address many of the issues relevant to the complexities of patient care by failing to integrate other nonstatistical forms of medical information, such as patient-specific issues and professional experience.12
Is EBM Evidence-Based?
The third theme offered by critics is perhaps the most ironic. EBM itself is not evidence-based, as it fails to meet its own empiric tests of efficacy. EBM is based on the notion that clinicians armed with the best available information will bring about an improvement in the quality of medical care. Despite the paucity of data to support the impact of EBM, many of its proponents remain firmly committed to this belief. Goodman, for example, suggested that to practice medicine in a manner inconsistent with EBM is unethical.13 To date, however, there remains no convincing direct evidence that demonstrates that care provided by practitioners whose practice is based on data isolated to applied health care research differ meaningfully from care given by those who rely on basic knowledge and their own clinical experience. Realistically, it is impossible to completely separate these two groups, as the complexity of patient care requires physicians to make use of many forms of information. So it seems that EBM suffers the same fate as so many issues in medicine that evade simple assessment by a randomized, controlled trial, and by virtue of that will likely not be supported by high-grade evidence.9
Impact on the Individual Patient
There may be some limitations of the effect of EBM on the individual patient due to both patient- and disease-related variables. Individual patient circumstances, stresses, and values will inevitably vary. And although disease-specific data will be available for the most common disease entities, gaps in reasonable data will continue to exist for varied subpopulations of patients, uncommon diseases, and disease variants.14 The clinical judgment provided by experienced practitioners will always be called on to resolve the complexities that exist in the setting of patient care.
Finally, loss of autonomy as it relates to the doctor-patient relationship by reducing the patients’ right to select treatments best suited for their individual circumstances is criticized by some as a potential flaw in EBM. This argument relates directly and indirectly to patient access to health care. EBM could be manipulated for use as a cost-cutting tool by preventing access to procedures or medications that are not proven to the standards of Grade A or B evidence. Further, EBM could increase the cost of health care by creating standards that require higher levels of proof before interventions would be available to the patient. To date, the true impact of EBM on cost and access to health care remains unknown.6,7
At its inception in 1992, EBM may have been seen as a challenge to, if not replacement for, traditional medicine. Evidence that was adherent to a narrow set of definitional standards was pitted against clinical judgment and physician experience, creating what appeared initially to be an adversarial interaction. Critiques, both practical and philosophical, have served to diminish the absolute differences between EBM and traditional medicine in lieu of a more balanced view of the role evidence-based medicine can play as a tool in the practice of medicine. Interestingly, none of the critics cited in this editorial suggest that high-quality evidence should be ignored in the overall context of clinical care. Instead, each argues that EBM should serve as one of many useful tools to help physicians address the varied and complex challenges of patient care.