|Home | About | Journals | Submit | Contact Us | Français|
Control and the therapeutic trial: rhetoric and experimentation in Britain, 1918–1948, Wellcome Series in the History of Medicine, Clio Medica 82, Amsterdam and New York, Rodopi, 2006, pp. 221, 46.00 (hardback 978-90-420-2273-7).
Featured on the cover of this book is an advertising poster for Hale’s Life Tonic, “for that ‘tired-out’ condition” mounted behind a painted image of a turn-of-the-century medical laboratory, with its acutely attentive researchers stationed at their instruments. Their life tonic is “control”, not the reality but the word, and bolstered by its enchantment they go boldly forth to slay the dragon of traditionalism. Therapeutic research in the three decades covered by this book marched under its banner, or so argues Martin Edwards, who brings an analysis of rhetoric to the history of the clinical trial in Britain. He is a bit ambivalent about the function of rhetoric of science, for while he understands well its legitimacy in any endeavour to persuade, often it seems to be accompanied by a tacit “mere”. The “control” that was exalted had no specifiable meaning, he argues, and a “‘controlled trial’ did not designate a single methodology, but signified approval for a trial conducted under the proper supervision and regulation of the M[edical] R[esearch] C[ouncil] and which should therefore, by implication, be regarded as trustworthy and reliable” (p. 176). Mere practitioners of medicine, though they sometimes experimented, were always vulnerable to the charge of inadequate control. Deprived of this tonic, they succumb to low spirits and brain fag.
This book is based principally on five episodes or case studies, sandwiched between an introduction and conclusion. The narrative template of the first four cases involves an illegitimate victory of illusory or meaningless “control” over alternative conceptions of medical expertise. In the first two of these, the MRC victory is won over faddish therapies of the 1920s: the treatment of diabetes by a diet of raw animal pancreas, and medical deployment of the healing power of light for a variety of ailments. The next two chapters concern tests of new therapies that were favoured by laboratory medicine: first serum therapy for pneumonia, then influenza immunization. Here Edwards is better able to bring out ambiguities, since the researchers themselves had to argue that conditions were somehow never quite right to make visible the effectiveness of their potions. His final case is a struggle between two versions of the medical experiment, one advocated by Almroth Wright, sometime opponent of statistics, and the other by Austin Bradford Hill, patron saint of the randomized clinical trial. Here Edwards speaks rather of semantics than of rhetoric, and interprets the triumph of statistics as a victory of clear terminology based on familiar English words over Wright’s commitment to a technical vocabulary of teeming neologisms that seem as barbarous to us as to his contemporaries.
Edwards does not succeed in showing that “control” was meaningless, though he does illustrate a fundamental shift of sense over the three decades covered by this study. The MRC in the 1920s did not demand control groups, but rather wanted as much as possible to hold all variables constant apart from the one under investigation. His charge depends on their failure to articulate a definition in the (predominantly) very short articles that make up their side of the disputation, and he does not look for an explicit doctrine in other sources such as textbooks. Advocates of (controlled) randomized trials, especially Hill, were at pains to expound their methodology. Edwards draws on a generation of scholarship, including several well-known but unpublished dissertations, to show how much these rationales ignored, and how differently the “gold standard” methodology functions in real life from the ideal. He also gives many examples of incoherent journalistic invocations of “control”, offering little beyond admiration for those modern types who know how to take charge and leave nothing to chance but randomization itself.