The Santayana Syndrome I: Errors in Getting and Interpreting Evidence

THE SANTAYANA SYNDROME I: ERRORS IN GETTING AND INTERPRETING EVIDENCE ALVAN R. FEINSTEIN* "Those who cannot remember the past are condemned to repeat it." This well-known comment by the historian George Santayana in 1922 is the eponymic source for the title ofthese two essays, but the basic idea is not new. Many centuries earlier, Cicero said, "To be ignorant ofwhat occurred before you were born is to remain always a child." For persons working in the field of health, the history of medicine is particularly worth learning for its combination of triumphs and blunders. The triumphs are generally well known. They include the technologic advances (mainly in the 19th and 20th century) that led to a sanitary public health environment for water, food, and sewage, and that have prevented or cured many infectious diseases. Other technologic advances have also dramatically cured or alleviated many non-infectious diseases, and have contributed to an almost doubled general life expectancy during the past 100 years. The blunders are generally not well known. At any era in the long history of medicine, however, most of the ideas about etiology and therapy of human illness were either flagrantly wrong or soon superseded by better ideas. As in any other era of medicine, the contemporary technology and data today are better than ever before, but current physicians are no more intelligent than our predecessors, andjust as likely to be immersed in doctrines that may later be shown to be erroneous. Mindful of this professional heritage and susceptibility, physicians are constantly challenged to prevent or to find and correct the contemporary errors. The "errors" to be considered in the two essays are ideas that come from wrong data or wrong reasoning, rather than foolish actions caused by negligence. The latter errors—such as amputating the wrong leg, removing the wrong lung, or giving a drug in toxic overdosage—occur when *Departments of Medicine and Epidemiology, Yale University School of Medicine, P.O. Box 208025, New Haven, CT 06520.© 1997 by The University of Chicago. All rights reserved. 0031-5982/97/4101-1035$01.00 Perspectives in Biology and Medicine, 41, 1 ¦ Autumn 1997 | 45 doctors know the right thing to do, but fail to do it because of careless lapses in simple, ordinary precautions. The distressing consequences of those lapses are usually promptly recognized and often receive extensive publicity, but the blunders themselves can readily be prevented and are not an acceptable part of routine medical practice. The fallacies under consideration now are ideas that have usually been well approved and maintained by the contemporary "establishment." and widely accepted as "guidelines" for practice in different eras of medicine. After the fallacies and their repetitions are presented, the discussion will turn to the potential role of medical history in preventing the Santayana Syndrome that occurs when past errors are repeated because they were inadequately identified, taught, or learned. The errors discussed in this first essay arise from problems in getting and interpreting evidence. The evidence may be [1] wrong, [2] correct but wrongly interpreted, and [3] made wrong by wishful thinking. Wrong Evidence The use of quantitative measurement in medicine is only about two centuries old, beginning with advances of the late 18th century and thereafter in chronometry, chemistry, thermometry, and standardized systems of weights and measures. Many users of modern statistical evidence do not realize that the famous "normal" (or Gaussian) curve was promulgated by Carl Gauss in the 19th century to deal with the new challenges presented by technologic problems of "observer variability." WTien disagreement occurred in repeated measures of the same entity, Gauss decided that the mean would be the "correct" value. Actual errors, rather than merely technical variations, have been a common phenomenon in the measurement of diverse entities. Older clinicians today can recall that Benedict's reagent was frequently used to identify glucose , but also gave positive results for galactose and other non-glucose reducing substances. Writh technical improvements, errors have been identified in many other measurements throughout the years. That errors can still occur with long-lasting consequences is seldom considered, however, for apparently simple measurements. For example, despite excellent sphygmomanometers and laboratory chemical procedures, wrong results regularly occur today for...