An Introduction to Clinical Research in Psychiatry

This is a book that needed to be written. Mental health clinicians and trainees often have little exposure to the basic issues related to clinical research, such as hypothesis testing, trial design, sample selection, study bias, and the use of appropriate statistical procedures for particular sets of data. These issues often are not addressed, or at most are casually discussed, during training. Thus many clinicians do not have the background to critically evaluate reports in journals or presentations at scientific meetings.

Drs. Blazer and Hays set out to rectify this situation by developing a primer for the clinician and the trainee. They have largely succeeded, although the book asks the reader to spend the time required to learn a new lexicon and digest a series of new concepts.

Fortunately, the book is well written in a simple narrative style that is easy to follow. The volume grew out of a course on clinical research methods that the authors teach at Duke University School of Medicine, and its organization and format reflect that genesis. Each chapter deals with a particular topic of relevance, is well organized, and has "assignments" at the end—study questions, with full answers.

The overall focus is on the elements of clinical research—research that is patient based or population based. The authors provide an overview of epidemiological methods, constantly returning to relevant clinical questions. The book begins with a discussion of causal inference and the nature of the scientific method; the work of Popper and Koch is used as a base. Then, in a crucial chapter, the authors discuss how to read a study published in a scientific journal. They dissect the structure of a paper, from abstract to discussion. They help the reader get inside a paper and begin to see what should be included and what can be left out. Problems and flaws with published articles are discussed, as are examples of excellent, masterfully written articles.

In the middle section, on study design, the authors first define the language of clinical research: populations under study; samples and sampling frames; proportions, rates, and risks; incidence and prevalence; risk ratios; and odds ratios. Following that is a chapter on single-subject studies, to help the reader understand the strengths of various single-subject designs, and one on descriptive studies. Examples of surveys of psychiatric disorders are presented. Next, the authors delineate the methods employed in epidemiological studies, including sample identification and community interviewing, as well as the use of longitudinal and case-control studies. The section ends with an in-depth discussion of the design and implementation of a classic clinical trial, including issues of protection of human subjects, sample recruitment, and use of comparison groups.

The final section, on analysis and interpretation of results, begins with a discussion of the use of diagnostic tests and screening instruments. The concepts of sensitivity and specificity are addressed, as is the application of screening tests to community samples. The next chapter reviews methods employed in population genetics studies and the relationship of these studies to molecular genetics investigations; concepts of heritability and methods for determining heritability are reviewed. The final chapter provides an overview of analytic strategies for clinical and community research data.

No doubt, An Introduction to Clinical Research in Psychiatry is a demanding book. As a text for an introductory course on research methods, it would work beautifully. As a stand-alone book, it would be more of a challenge. But it is well worth reading. Anyone who spends the time to read the book and answer the study questions will obtain a solid grounding for reading the literature and for beginning to work in clinical research.