Biostatistics 101: Understanding Data
David Etzioni, MD, MSHS; Maher A Abbas, MD, FACS, FASCRS
Summer 2009 - Volume 13 Number 3
Kaiser Permanente (KP) is a leader in health care delivery and provides care for millions of Americans in several regions and states, including Northern and Southern California, Colorado, Georgia, Hawaii, Ohio, the Northwest, and the Mid-Atlantic States. The volume of clinical care rendered every year throughout the organization presents a great opportunity for research and innovations. In recognition of the importance of research, the Kaiser Foundation Research Institute was created in 1958 to administer and support research within KP at a national and regional level. High-quality innovative translational research is performed every year, in the form of randomized clinical studies, epidemiologic research, retrospective databases review, and health care policy research. Supported by an electronic medical record and computerized databases, KP is well-suited to provide the scientific community with a wealth of data on outcome of interventions. In 2007, there were approximately 2800 active studies, all approved by institutional review boards, being conducted nationally. That same year, reports on 571 studies were published by KP physicians and scientists in prestigious medical, surgical, and scientific journals, including The Permanente Journal.
Because of the size of the KP population, KP research studies often contain a large number of study subjects. Such projects and endeavors can generate complex data and results that require analysis to demonstrate the effect of therapies and interventions, to establish the efficacy or limitation of treatments, and to prove or to refute a scientific hypothesis. An understanding of biostatistics is critical to the researcher investigating clinical questions. Equally important is an appreciation of statistics by the reader and interpreter of published studies. As with all fields of scientific endeavor, statistics encompasses a rich jargon that is necessary to abbreviate and refer to underlying concepts. In this article, the first of a three-part series on statistics for clinicians, we begin with an overview of how statistics can and should be used to describe data.