Half-life of truth in surgical literature

Abstract

The hypothetico-deductive model of Karl Popper contends that “An assertion is true if it corresponds to, or agrees with, the facts”. Because “the facts” change over time, truth is relative. In this view of events, science progresses via a series of theories (paradigms) that are held to be true until they are replaced by a better approximation of reality. It follows that some “truths” will last longer than others. We decided to estimate the half-life of dogma relating to the practice of surgery. The journal Surgery Gynecology and Obstetrics had a section entitled International Surgical Abstracts that concluded at the end of 1994. We obtained copies of all of the abstracts for the even numbered months for each fifth year after 1935. The first 20 abstracts about general surgery from each of the 13 review periods were selected for study. The one sentence from each abstract that best summarised the conclusion was copied to a database. Editing of these sentences was restricted to the rephrasing of outdated terminology and the elimination of redundant words. The form, which contained the 260 selected sentences in a random order, was assessed by seven general surgeons. They were asked to mark each question as being either true or false. We explored the relation between time and the number of positive responses from the questionnaire using scatter plots and regression analyses. The linear model detailed in the figure was found to be most appropriate (R=0·86, p<0·001). It suggests that the rate of loss of truth is 0·75% per year and that the estimated half-life of truth for clinical statements in the surgical literature is 45 years. This seems appropriate; 45 years ago it was suggested that: prefrontal lobotomy usefully altered patients’ reactions so that “no anxiety, fear, or concern over their impending death from cancer was manifest”; “in primary malignant hypertension the malignant phase may disappear” after lumbodorsal sympathectomy; and, the detection of a gastric ulcer was “a strong indication for immediate operation”. Extrapolation of the regression line can be used to hypothesise that the current era of surgery extends between 1904 and 2038. From an historical perspective, 1904 is a reasonable estimate of the commencement of the era of “open” surgery. Gauze masks and rubber gloves had just been introduced into operating theatres and abdominal surgery gained public respectability in the UK in 1902 when Edward VII’s coronation was delayed by the need to drain his appendiceal abscess. More contentious is extrapolation of the regression line to predict the future. However, a nexus point for the present era of surgery in the second quartile of the next century is consistent with the current trend towards the ascendancy of minimally invasive procedures and medical treatments over open surgery.