Evaluating Evidence

PROLOGUE: Notwithstanding the epidemic of hype that surroundsand obscures the concept of evidence-based medicine (EBM), increasedefforts to systematically collect and disseminate clinical researchresults to practitioners are a necessary and appropriate responseto the contemporary challenge of exploding biomedical knowledge.But the struggle to tame this unruly frontier is a contentiousbusiness. Basic agreement about the proper organization of itsboundaries and infrastructure remains to be achieved. Some partiesto the medical enterprise question the fundamental thrust ofEBM and suspect it to be a Trojan horse concealing an intentto industrialize and cheapen medicine. For others, that debateis over, and the only question is how more and more EBM canbe implemented.

But even among those who agree that a more robust and better-organizedflow of scientific information is the key to improved safety,quality, and efficiency, many questions remain about how toevaluate, grade, prioritize, and use clinical research resultsin practice. With thousands of new studies appearing every month,how can clinicians weed out specious or misleading findings?How are credible studies with nuanced differences to be organizedand applied? What innovative approaches to clinical trial designand analysis show the most promise for bridging the gaps betweenbench and bedside?

The following papers present a variety of approaches to sortingand evaluating evidence. Earl Steinberg and Bryan Luce reviewthe many diverse contextual factors that may enter into theevaluation and application of clinical research evidence. Theyconclude that despite the availability of rigorous methods forweighing evidence about the effects of medical interventions,these methods are not always applied and interpreted correctly,and not all "evidence-based" practices have equal claim to validity.Karl Claxton and colleagues propose a pragmatic framework inwhich evidence of variable strength and quality can be set toinform rational decision making. Their approach, illustratedby the British National Institute for Clinical Excellence (NICE),centers on "value-of-information analysis."

It is a paradox for policymakers that two different medicaljudgments may conflict even though both have plausible claimsto be evidence based. David Atkins and colleagues explain howsuch conflicts may be understood not as a confounding of EBMbut as the application of evidence in the tacit context of differingassumptions about underlying values and purposes. Finally, DanFox reviews the historical development of the discipline ofsystematic research synthesis, which has been the cornerstoneof the EBM movement in recent years. Fox uses the example ofthe Drug Evaluation and Review Project (DERP) in Oregon to explainthe promise and pitfalls of systematic reviews. Following thesepapers are three Perspectives on evaluating evidence by MarkHelfand, Steve Teutsch and colleagues, and Dan Mendelson andTanisha Carino.

What's this?

Home

Current Issue