In recent decades, academic physicians and health policy analysts have espoused the use of randomized experiments (RCTs) for obtaining reliable, scientific data about the benefits of medical treatments. Since the early 1900′s, the controlled therapeutic experiment has played a dual role as an arbiter of scientific validity and a tool for regulating medical practice. Such experiments have been the chosen instrument of a small cadre of medical researchers who have sought to constrain and direct the average practitioner’s use of medical treatments. How and why these reformers thought improvements in the quality of therapeutic research would naturally benefit clinical practice, and the steps they took to bring these improvements about, are the subjects of this thesis.
The thesis examines the activities and scientific practice of professional reformers on the Council of Pharmacy and Chemistry of the American Medical Association (AMA) in the 1910’s and 1920’s; the influence of these reformers on the Federal Food, Drug and Cosmetic Act of 1938; and the intellectual and practical problems faced by the organizers and government sponsors of controlled therapeutic trials up to the 1950’s. I then examine postwar developments, including the roles of statistics and statisticians in experimental design, controversies over the design and interpretation of RCT’s and current proposals to use RCTs as a guide in allocating medical resources.
My argument is first, that our public policies and practices for regulating therapeutic drugs during the first half of this century were developed and shaped by private institutions, most notably the AMA’s Council on Pharmacy and Chemistry, and the community of academic physicians represented in the National Research Council’s Division of Medical Science. Second, that in focusing their efforts on the scientific problems of improving the evidence on which judgements about the risks and benefits of medical treatments are based, therapeutic reformers have avoided the more intractable political problems presented by attempts to regulate or manage the therapeutic decisions of practising “good” clinical decision is as much a problem for political theory as for statistics and/or “clinical” judgement.
The problem of reforming the therapeutic practices is treated here as an instance of the more general problem of the relation between science and democracy. In the concluding chapter, I examine contemporary proposals to rely on randomized experiments in determining public policies towards the use of the medical practices and technologies. I argue that if we are to regard the use of RCTs in evaluating medical technologies as more than an administrative convenience, the theoretical claims that RCTs provide and authoritative basis for evaluating and allocating medical goods must be assessed. I analyze the relation between method and science implicit in contemporary proposals for therapeutic reform, and argue that a purely methodological theory of science, as is offered by advocates of RCTs, cannot provide a basis for either scientific judgments or public policy. I conclude with a series of arguments for a more participatory, more public, approach to setting policies towards medical practices, as opposed to relying on more corporatist arrangements which delegate authority and responsibility to the medical community and the private sector.
The methodology of the thesis is largely historical: I have used manuscript collections to identify the individuals and organizations active in therapeutic reform, their activities and their ideology. Collections consulted include the archives of the AMA, records of the Food and Drug Administration in the National Archives and on deposit at the Federal Records Center, records of the National Research Council, the National Institutes of Health, and a variety of private manuscript collections. These materials have been supplemented with interviews and with analysis of the published materials.