BOOK REVIEWS Edited by H. Stanley Thompson, M.D.
The last section of the book is devoted to ethics in clinical research, a subject rarely covered in a book on methodology. While the values expressed in the authors' approach to human subjects are common to all ethical professionals, they must be stated and restated. As new technologies and areas of research emerge, we must continue to apply tests of ethical behavior to their investigation. While this book clearly explains the methodology of clinical research, it does not effectively teach the underlying methods. There are no statistical tables or recipes for performing statistical tests. The book makes no attempt to serve as a reference book in either epidemiology or biostatistics. Although there is no shortage of reference books on either subject, additional technical detail in the appendix and some rudimentary statistical tables would have increased this book's readability. This book does, however, introduce a reader to a highly technical subject in an engaging and enjoyable manner.
Research for Health Professionals: Design, Analysis, and Ethics. By Robert Proulx Heaney and Charles J. Dougherty. Ames, Iowa State University Press, 1988. 282 pages, index, illustrated. $39.95
Reviewed by JONATHAN C. JAVITT Baltimore, Maryland Reading Drs. Heaney and Dougherty's new book "Research for Health Professionals: Design, Analysis, and Ethics" is more like having a conversation with two friends than reading a book on research methodology. I will readily admit to a prejudice that most books on statistics, epidemiology, and clinical research are tough reading at best. This book is approachable, well written, and enjoyable. That said, it is important to define its potential role on an ophthalmologist's bookshelf. This is a book on the why's and wherefore's of clinical research. It brings together epidemiology and biostatistics as integral and essential methodologies and avoids the trap of getting overly involved in the technical aspects of either. On rare occasions, this can be frustrating. For example, after spending three pages discussing the importance of choosing an adequate sample size, the authors tell us to look up sample sizes in a table (presumably published elsewhere). An enhanced appendix with a few more basic formulas would be helpful. The authors introduce us to basic concepts of probability and inference using everyday examples such as coin-flipping, card shuffling, and the like. They then proceed to apply these principles to the design and execution of clinical investigation. They lucidly explain the concepts of sampling, randomization, and hypothesis testing. At every juncture, they point out the most frequent errors made in clinical research, along with ways to avoid them. The third section of the book is devoted to the application of statistics. The authors have replaced formidable equations with easily understood English. An example is their explanation of parametric vs nonparametric statistical methods, a concept that is generally buried in a morass of incomprehensible equations. A chapter is devoted to common errors in the application of medical statistics. The simple concepts it encapsulates may save an author from embarrassment in the peer review process.
Laser Photocoagulation of Retinal Disease. By Kurt A. Gitter, Howard Schatz, Lawrence A. YannUZZi, and H. Richard McDonald. San Francisco, Pacific Medical Press, 1988. 281 pages, illustrated. $65
Reviewed by RICHARD F. DREYER Portland, Oregon Contained in this volume are the proceedings of the International Symposium on Laser Photocoagulation of the Macula, held in New Orleans in November 1986. Seventy-seven experienced laser specialists contributed. Separate sections deal with diabetic retinopathy and macular degeneration. More unusual conditions are briefly reviewed, including laser treatment for malignant melanoma, retinoblastoma, and retinal angiomas. Such a symposium with a large number of contributors provides a forum for work in progress, allows for the presentation of new techniques that mayor may not prove useful, and provides opportunity for discussion. This book both benefits and suffers from this format. One of the benefits is that a broad range of 310