This book is a worthwhile survey of the philosophy of Evidence-Based Medicine (EBM). EBM is, roughly speaking, the idea that all clinical practice ought to comply with current orthodoxies in clinical epidemiology. Philosophers are having to play catch-up with clinicians in this area, and so it is very good to know that creative and serious minds like Goodman's are on the job. Goodman understands the history and sociology of EBM very well, and dissects it admirably, carving it accurately at the joints.
Goodman covers a fascinating range of topics, centred on the following: the meaning of EBM (probably the most controversial topic of all); the applied epistemology of synthesizing disparate pieces of research evidence; the ethical implications of public data on the Web; the rigidity or otherwise of clinical treatment guidelines; and the effects of scientific uncertainty on public health policy. All this is supported by a good, up-to-date bibliography. I particularly like the many areas in which Goodman points out types of fallibility in clinical science which had previously escaped my attention and which have likewise escaped the attention of most proponents of EBM. The only important topics (to my mind) that Goodman bypasses are controversies over statistical techniques which threaten the validity of the methods recommended by EBM, and the ethics of EBM in the Third World and other poor and powerless jurisdictions.
This is the sort of book that attempts to be more or less encyclopaedic in its area. It will be enjoyable and useful for people who are looking for a field guide to a large number of philosophical topics relevant to EBM, grouped in a sensible and helpful way, and who are happy to be told not much about each of them. From what I can gather, these people are the book's intended audience. It also traverses enough territory to provide opportunities for many entertaining bons mots, some of which stand on their own, not requiring support from detailed argument.
On the other hand, as an attempt not only to provide a critical introduction to EBM but also to understand EBM in any philosophical depth, this survey is not so successful. The short length of Goodman's book (I estimate less than 50,000 words, and chatty words at that; this is no Tractatus), combined with its sheer scope, combined with the fact that many of his topics are treated by quoting from a wide variety of other authors (who have a wide variety of depth of insight) rather than providing us with Goodman's own synthesis, leads to an inability to convince me, and, I suspect, any philosopher, that he has gotten to the bottom of any of his topics. So, terrific though this book is for some audiences, its appeal to philosophers of science is bound to be limited as a result of its rather diffuse focus.
Goodman's attempt to cover all the bases leads him down hundreds of short and unenlightening dead ends. Some of the back alleys turn out to be through-roads, but most of them don't. For example, it will not be helpful to philosophers to read a couple of short paragraphs about the distinction between knowledge and skill (14) only to come to the conclusion that “this distinction can propel us into vast and unproductive disputes” (15). This conclusion is merely asserted, not argued for, and no further conclusions are drawn from it. I personally find this topic interesting and agree with Goodman's conclusion, so if he had analyzed the topic in any detail I would have enjoyed reading about it; but as it was I found this passage unhelpful, along with many similar ones.
To make all my above remarks more concrete, consider one theme from the book (more or less at random): the suite of problems associated with the increasing interdisciplinarity of medicine. Of course medicine has always been interdisciplinary, but several recent changes—such as the new importance of biostatistics, bioethics, and law in recent years (and, I hope, the new importance of sociology in future years)—have made the traditional medical education, with its extreme emphasis on biology, appear obviously inadequate to the leaders of EBM. A particular issue is ensuring that clinicians are competent in the disciplines that they have not traditionally been taught. This leads to problems, firstly because these are relatively unformed disciplines—the major questions in biostatistics may have been answered (though I don't think so), but in any case the answers are so contentious that many a prominent biostatistician would fail an exam set by a rival biostatistician from a different school—and secondly because these new aspects of medicine are news to clinicians who completed their formal educations some time ago or at universities that were behind the times. Goodman's discussion of this problem is not especially deep, and reaches not much of a conclusion. On the other hand, it does map out the territory with exemplary clarity.
Some few of Goodman's topics offer more than this. A nice example is his discussion of the assertion that medicine is more of an art than a science, an assertion which Goodman argues convincingly (95 ff.) to be no more than wishful thinking.
