Statistics from Altmetric.com
The issue of trial design is increasingly recognised as central to the evaluation of the benefit from any healthcare intervention. The author describes the book as an introductory guide aimed primarily at clinicians, young researchers, research staff and trainees in any health profession.
My personal view, admittedly, in the author's own words (p 81) “idiosyncratic, ... informal and subjective”, is that while I fully endorse all that is said, these are not quite the right contents to target at this readership. In particular, chapter 4 deals with assessing the quality of randomised controlled trials, an important issue for all readers, but perhaps not to such depth; the section on developing a formal instrument to assess trial quality will be of direct concern to few.
Conversely, there is much that has not been included, and that could profitably be enlarged upon. The author admits that the book is not meant to be a statistics text, and refers readers to Altman's excellent book, Practical statistics for medical research (Chapman & Hall, 1991). Nor should it be, but it would be worth enlarging a little on the key issues of sample size and power, and the preference for point and interval estimates of effect size rather than hypothesis tests, and cross-referencing another booklet published by BMJ Books, Statistics with confidence, with accompanying software, second editions of which are to appear this year. The author refers to the recognised need to make the most of existing knowledge, leading to a strong evidence-based medicine emphasis, though the other consequence of this, health services research, is not elaborated on.
The book would be easier to assimilate with a few more examples. This is done very well in chapter 3, in which the issues are developed in the context of a new drug to delay disease progression in multiple sclerosis. This approach could also have been profitably applied elsewhere.
I will find this book very useful as a source of material for my teaching, refereeing and research collaboration. But it is not exactly what I would have written as an introduction to the subject for clinical researchers.