The ascendancy of EBM has been accompanied by a greater awareness of its shortcomings. It is increasingly evident from the cost, length, and difficulty of performing RCTs that studies cannot be launched to address every question in medicine. Good evidence is often lacking in medicine. Epistomologists question the very notions of evidence and the suitability of current study designs and measurement tools to research the salient issues of concern to patients and others concerned with quality. Lack of evidence of effectiveness does not prove ineffectiveness, yet, in reaction to EBM, insurance companies and government often make this inference to justify decisions to withhold coverage or research support. The unbridled enthusiasm for the evidence-based practice guideline of the early 1990s has been tempered by a more mature understanding of its limitations. Not many practice guidelines are developed well, and the implementation of flawed guidelines can cause harm. The seven-step process outlined earlier is slow, laborious, and expensive (sometimes costing hundreds of thousands of dollars). Moreover, there is little evidence that either the rigor of the methods or the guidelines themselves have a meaningful effect on practice behavior or patient outcomes. To the most cynical observers, the only consistent beneficiaries of guidelines are payers, who use guidelines with considerable success in reducing costs, lengths of stay, and utilization rates. Even ardent advocates of guidelines acknowledge the evidence that disseminating reviews and recommendations, by itself, fails to motivate clinicians to increase delivery of effective interventions and to abandon ineffective ones. This absence of response has stimulated a closer look at the barriers to behavior change and the design of thoughtful implementation strategies that begin with, but reach beyond, simple guidelines. Tools such as reminder systems, standing orders, academic detailing, peer review and audit, feedback, and health system changes recognize that knowing what to do is only one piece of an increasingly complex puzzle. The competitive marketplace of managed health care has added new economic influences on clinician behavior but is also fueling private-sector interest in good research. Patients, clinicians, and policy makers will continue to seek better data concerning what works in medicine and what does not.