Skip to Main Content

The quantity and diversity of scientific publications are vast and rapidly expanding. Yet, the quality of the underlying methodology is highly variable and many studies have poor study design or analysis. While editors at traditional journals invest substantial efforts to help ensure published studies are performed and interpreted correctly, readers of those journals will be better equipped to apply study results to their individual clinical practice if they appreciate the strengths and weaknesses of the underlying research methodology. With the increasing popularity of preprint servers, clinicians will read scientific reports prior to rigorous peer review, which requires the ability to independently evaluate the design, analysis, and interpretation of clinical studies. Now, more than ever, clinicians must have sufficient understanding of study methodology to judge if what they are reading is likely to be valid and applicable to the patients they treat.

Statistics is generally not a very popular course in medical school; thus, many clinicians do not have a firm understanding of traditional study design and statistical analysis. Many medical statistics courses emphasize mathematics along with myriad assumptions and rules that do not resonate with clinicians who pursued medical school because of their love of biologic science and interest in helping others, rather than an interest in mathematics. Further, there are many new statistical methods that are increasingly important in clinical research that clinicians may have never seen before.

The JAMA Guide to Statistics and Methods was created to help clinicians overcome these challenges. Research methods are described in plain English without reliance on equations and without any assumption that the reader understands the field of statistics. We distinguish between statistics and methods: statistics are mathematical approaches to describing collections of data and associated uncertainty, whereas methods refer to how a study was designed or some other aspect of how a study was organized and conducted. Each article explains how a statistical or methodologic concept, or the results of a statistical analysis, should be applied or interpreted in an example research article published in JAMA or JAMA Network journals, along with the limitations associated with the data and the methodology used to evaluate them. The intent of these guides is to enable clinicians to better understand research findings so that they can competently assess if study results are reliable enough to adopt them into clinical practice.

This book is organized into major sections: Interventional Studies and Observational Studies. Each of these has 4 subsections. For Interventional Studies these subsections are (1) Trial Strategy and Design, (2) Enrollment, Allocation of Treatment, and Ethics, (3) Measurement of Outcomes, Analyses, and Interpretation of Results, and (4) Application of Results. The subsections for Observational Studies are similar: (1) Study Strategy and Design, (2) Assessment of Risk Factors and Exposures, (3) Measurement of Outcomes, Analyses, and Interpretation of Results, and (4) Application of Results. A third section provides guides for interpreting studies of large surgical data sets to enable clinicians to better understand these types of studies.

We thank the contributors to this series who represent leaders in clinical research, methods, and statistics. They are listed on the following pages.

The JAMA Guide to Statistics and Methods is not meant to be a comprehensive overview of all possible methodology used in medical research. Rather, by basing topics on recently published articles, methods emphasized are those commonly in use. For clinicians negotiating the ever-expanding medical literature, these topics will be the ones that they are likely to see and must understand to determine the relevancy of research findings to their clinical practice.

Edward H. Livingston, MD

Roger J. Lewis, MD, PhD

Pop-up div Successfully Displayed

This div only appears when the trigger link is hovered over. Otherwise it is hidden from view.