Critical Appraisal of Medical Literature

There has been an enormous number of publications on medical statistics and critical appraisal in recent years. This is in keeping with the recent trend in the application of evidence-based medicine (EBM) in clinical practice and clinical audit, as well as professional examination.

We are often confused by the number of EBM textbooks that are written by people from different professional backgrounds. To most psychiatrists, books geared towards polishing critical appraisal skills that are suff icient for examination and continuous self education pur poses, perhaps written by psychiatrists, would be most relevant and appealing. Existing books on EBM usually adopt various approaches. Some of them start from research questions, research methodology, and hypothesis testing while some adopt the ‘top-down’ approach by appraising the study results yielded by different research methodologies. This book, however, is written by a psychiatrist who adopts both approaches.

The book is divided into 3 parts. Part I is devoted to the discussion of the basic concepts of EBM, in terms of probability, variable and measurement, distribution, hypothesis testing, research methodology, bias, and errors as well as epidemiology. Part II focuses on the approaches to critical appraisal of a wide range of research papers. Part III deals with the more sophisticated statistical concepts of multivariate analysis and non-parametric tests.

All topics are organised into short, concise chapters that can be read in less than 15 minutes. The theories are illustrated both conceptually and mathematically and narrated in lucid English. The author has included summaries in note for m at the end of each chapter that are comprehensive yet concise enough to allow one to skip the main text if little time is available for reading. This is particularly important for examination candidates who look for materials to help with multiple choice examinations and short questions.

Part II of this book is the most relevant section to the critical review paper of the Member of the Royal College of Psychiatry (MRCPsych) Part II Examination. The section starts with a few chapters giving the general framework of critical appraisal technique, which is then consistently applied to the appraisal of different research designs in the subsequent chapters.

This section encompasses papers on treatment, diagnosis, aetiology, literature review and meta-analysis, audit, clinical guidelines, and qualitative research. Checklists of items to consider in appraising different research methodologies are compiled in a summary chapter at the end of Part II providing a quick tool of reference.

Part III endeavours to cover the various statistical techniques, including the z-test, t-test, analysis of variance, non-parametric tests, correlation, regression and prediction, and multivariate analysis. The coverage is adequate for most clinicians and potential MRCPsych examination candidates. On the other hand, in-depth illustrations of the mathematical steps and tabulation processes of statistical analysis are lacking, which makes it diff icult for beginners to follow. The author has, however, made it quite clear that it was not his aim to teach the computational technique in-depth, hence this part of the book cannot replace standard statistical textbooks and manuals.

In general, this book serves to reach the general audience with little or no background in research methodology and EBM. It is of great relevance to MRCPsych examination candidates for a quick grasp of the essential approaches to critical appraisal. However, exercises for taking the Membership examination are lacking. Therefore, examination-oriented readers might need to refer to other exercise books on critical appraisal.

Dr Sandra SM Chan
Assistant Professor
Department of Psychiatry
The Chinese University of Hong Kong
Hong Kong, China

View My Stats