The economic evolution of American health care: from Marcus Welby to managed care

of the profession in deciding how drugs were to be evaluated. If the ideal of a profession united by science fell by the wayside, so too did another key element of reform ideology; the trust placed in the high moral character of individual clinicians to ensure the integrity of observations on the therapeutic effects of drugs. Marks argues that during and after the Second World War this trust began to fade. It was replaced at the urging of statisticians and others, with a reliance on formal statistical methods-notably the double-blind, randomized, controlled clinical trial as means of assessing and improving therapeutic knowledge. A faith in method as the motor of reform superseded a faith in the moral qualities of individuals. Marks is not the first to trace a shift from a trust in people to a trust in numbers, but his may be the strongest historical voice arguing for the incompleteness of this change. As advocates of clinical trials constantly remind us today, the clinical trial has not permeated all areas of medical research, and most therapeutic practices have not been subject to the probe of a trial. It is also true that experts disagree on the value of particular designs for clinical trials, and on how to explain results. This then is the story of an incomplete revolution. It is grounded in detailed case studies: (before the Second World War) the Cooperative Clinical Group's study of syphilis treatments, and the Commonwealth Fund's experiment with serum treatment for pneumonia; (during and after the War) the National Research Council's penicillin investigations, and the Veteran Administration/Public Health Service investigations of streptomycin; and (in the 1960s) the National Institutes of Health planned Diet-Heart Study, and the University Group Diabetes Program study of tolbutamide. Together these examples help to flesh out the story of this partial revolution. They also provide an insight into the tensions between researchers and general practitioners, and the impact of patient compliance on experimental design. It is possible to quibble that the story tells us less than we might wish to know about the nuts-and-bolts of therapeutic testing, and that Marks' assertion that this is a quintessentially American tale lacks a comparative perspective to nail the point home. These caveats aside, Marks has produced an important account of twentieth-century clinical medicine; conceptually sophisticated, and supported by a rich lode of footnotes. The cheaper edition of this book will ensure that future students can afford to mine the latter at their leisure.