“in the Balance”: Weighing the Evidence

In their classic text on how to analyze and revise expository prose, The Reader Over Your Shoulder [1], Graves and Hodge describe the “natural arrangement of ideas in critical argument”: A question or problem is stated; evidence supporting, or conflicting with, a tentative answer or solution is presented; the relative strengths of all evidence are weighed; and the answer most strongly supported is delivered. They were writing about argument of any kind—a debate on foreign policy, criticism of a government's fiscal management, literary criticism—and not specifically about “argument” in science. Their notion of critical argument applies, however, just as well to resolving questions through the procedures of science and reporting the results of research [2]. The terms used for the elements of critical argument in science may differ from theirs, but the essence of their view is as correct for science as for other fields. The initial question or problem posed for a piece of research may be called “hypothesis” and the evidence may be called “results” and “the cited literature.” Regardless of the terms, the aim in science is to take all the available evidence possibly supporting or rejecting a hypothesis or potentially answering a question and to come to a conclusion.

A physician facing the question of the best treatment for a particular patient cannot take even a tiny fraction of the time needed to weigh all the available evidence bearing on the decision. The physician needs an answer and quickly. Most of the time, physicians must carry the answers around in their heads. Much of this information typically represents judgments reached by experts who have taken the time to weigh evidence from reported research. Their judgments are set forth in textbooks, in postgraduate courses, and in consultations. But medical journals also have the responsibility for transmitting expert judgments. Many of these judgments are those reached by authors of papers reporting their clinical research. But these judgments may not be unanimous among experts. When controversy exists, journals can additionally help by presenting analytic critical judgments about particular treatments in the form of editorials, review articles, meta-analyses and letters-to-the-editor. In these formats, however, only the judgment of the author is presented, and the reader may not be aware of dissenting judgments that are worth consideration before a treatment is accepted or rejected.

To help our readers develop their own conclusions on the advantages or disadvantages of a treatment about which experts differ, we are initiating a new section, “In the Balance.” The title is meant to represent the aspect of critical argument that Graves and Hodge call “the weighing of conflicting evidence.” The editors will select the topics to be discussed. Our plan is to present in this section, from time to time, dissenting views on questions of diagnosis and treatment, particularly questions arising in the care of patients with common, important problems.

The proper, efficient treatment of acute myocardial infarction is a question of major importance in internal medicine. “In the Balance” debuts in this issue with a pair of papers [3, 4] that present differing views on the results in the GUSTO trial of thrombolytic agents reported last year in The New England Journal of Medicine[5]. To help our readers develop their own critical skills, we asked Dr. David Sackett to consider issues arising from the GUSTO trial report using an approach that illustrates how physicians can think for themselves about trial reports. The resulting analysis [6] is presented in the editorial preceding mine.

We hope that the “In the Balance” papers and the editorial in this issue will help physicians reach their own conclusions about thrombolytic agents and will illustrate for them a method for judging critically the importance of conclusions reached in a clinical trial.

• Copyright ©2004 by the American College of Physicians