Understanding evidence-based medicine

Systematic reviews in health care: a practical guide. Paul P Glasziou, Les M Irwig, Christopher J Bain, Graham A Colditz. Melbourne: Cambridge University Press, 2001 (viii + 137 pp). ISBN 0 521 79962 7.

Why should you choose this book in what is a relatively crowded market? The authors are Australian experts and well qualified to write it. It is an introductory text, presumably for Master of Public Health students or mid-career clinicians attempting to come to grips with one of the foundation stones of evidence-based medicine. The flyleaf states “this is a book for those with an interest in synthesising healthcare research and for those studying for a degree in public health”. The book provides an excellent overview of the general methods of systematic reviews and is a useful primer on the topic, particularly the chapter on diagnosis and the discussion concerning heterogeneity. The second part addresses question-specific methods. It has some exercises at the end of each chapter, but would benefit from the inclusion of worked examples. Inevitably with a text of this size, the question arises: What was left out that should have been included? I think a more substantial treatment of the relationship between study results reported as proportions, odds ratios, relative risks and the number needed to treat (NNT) is warranted, with appropriate references (a passing reference is provided in the question section at the end of the chapter on interventions). The chapters on interventions, frequency and rate could benefit from the incorporation of “look-up” answers to the questions. The chapter on diagnosis should discuss the diagnostic odds ratio as a summary measure of the accuracy of a test. The answer to the question “Should I buy this book?” is “Yes”, if you want to get started and wish to move beyond the User’s guide series in JAMA. I would also advise downloading the Cochrane review handbook. Donald Campbell Clinical Epidemiologist Royal Melbourne Hospital, VIC NB: The reference (page 115 to the website for the Easy MA software) is incorrect (it should be www.spc.univ~lyon1.fr/mcu/easyma/).