When I was in high school I thought that my English classes were a complete waste of my highly valuable adolescent time. Of course, they turned out to be by far the most valuable education that I obtained in high school. Fortunately, I had some very good teachers. In college I took a lot of advanced math classes in preparation for a degree in physics. How I took a turn into medicine is a topic for another time. My math courses included topics like vector analysis, advanced calculus and linear differential equations, all of which I've completely forgotten. Somehow, however, I managed to avoid a basic statistics class. That's something I've regretted for many years.
Basic statistics is an invaluable tool that should be better understood by all medical authors (and readers). The topic is at the foundation of so-called "evidence-based medicine" a concept that is much on vogue in all of medicine. As health care dollars become more difficult to obtain, there will be an ever greater emphasis on providing diagnosis and treatment in the most time and cost efficient manner possible.
Against this background, Dr. Alberto Marchevsky, along with my friend and colleague, Dr. Mark Wick, have edited a concise multi-authored text entitled, "Evidence Based Pathology and Laboratory Medicine," published by Springer. The chapter authors are all well-known experts in their field. Chapter titles from the first half of the book include:
Biostatistics 101
Prognosis and Prediction in Anatomic Pathology
Power Analysis and Sample Sizes in Pathology Research
Meta-analysis: A Statistical Method to Integrate Information Provided from Different studies
Decision Analysis and Decision Support Sytems in Anatomic Pathology
...and many others.
The second half of the book is devoted to solutions offered to pathology and laboratory medicine using an evidence-based approach. Topics here include, among others:
Evidence-based approach to interpreting and integrating pathology literature.
Evaluating and reducing diagnostic errors using an evidence-based approach.
Evidence-based practices in applied immunohistochemistry.
Evidence-based pathology in molecular diagnostics.
As an example of the insights provided, have you ever wondered how many cases it would take to validate and refine our criteria for distinguishing an encapsulated follicular variant of papillary carcinoma from a follicular adenoma? Both, have extremely good prognoses, and because of this one would need approximately 120,000 patients followed for at least 10 years each to fully sort out the morphologic distinctions. Needless to say, this is unlikely to ever happen.
This book is an extremely well done, thought provoking series of treatises on how to interpret medical literature and improve pathology practices. Each chapter stands as an essentially independent topic, and the book can be "cherry picked" or read straight through. Since one of the editors and several of the authors are good friends, I might be accused of bias in recommending this book, but I really believe it is an excellent treatment of an extremely timely topic.
See if you agree, and please post your comments.