Essential Evidence-based Medicine

Yet another beginners' guide to evidence-based medicine (EBM) entering an overcrowded market ought to have something exceptional to offer. This book does. I liked Dr Meyer's approach to explaining the concepts and skills that frighten beginners. He uses plain language effectively (most of the time) and backs up the written explanations with helpful pictorial ones, which should appeal to doctors worried by mathematics. This approach marks the book out from many others. However, the style is at times too glib. ‘How should guidelines be developed? The process of guideline development should be evidence based’ (page 278). Those sentences, giving no surprises and even less insight to the reader, occur too frequently to be overlooked. The content is fairly comprehensive, including as it does some basic epidemiological tools as well as appraisal. However, the balance could be improved. Few teachers of EBM would consider the lengthy explanation of odds ratios was justified. It was disappointing to find that qualitative research — which is important — had been left out while space was devoted to decision analysis — which is unimportant in real practice, a fact indicated by the author himself. Apart from the last example, a good feature of this book is the amalgamation of EBM principles with recommendations for their use in clinical practice. The rational discussion on prescribing antibiotics for sore throat is a good example of such an amalgamation. In case you are wondering, yes, Dr Meyer does prescribe in some circumstances but more importantly he explains how he reached the decision to do so. The accompanying CD-ROM has a partially interactive programme containing worked examples of critical appraisal, calculations, and MCQs enabling the reader to practice and learn through correction. The book and CD-ROM combination provides an excellent instructional manual that marks this book out from several books whose accompanying electronic medium adds little value to the text. However, I feel that the package could be greatly improved by better editing. Apart from the style problem already mentioned, there are lapses where words are used sloppily. For example, the CD-ROM refers in one instance to the point estimate as the ‘actual difference’. It defines event rates as ‘the number of desired outcomes divided by the total number of all outcomes.’ Not if the outcome is death. What is more, the definitions for event rates in the book and CD-ROM are different. Beginners will find this package useful but hopefully the second edition will improve on the style.