Transactions of the Sixty-Third Annual Meeting of the South Atlantic Association of Obstetricians and Gynecologists
Understanding evidence-based medicine: A primer

Presented at the Sixty-third Annual Meeting of the South Atlantic Association of Obstetricians and Gynecologists, Hot Springs, Va, January 20-23, 2001.
https://doi.org/10.1067/mob.2001.116740Get rights and content

Abstract

Evidenced-based medicine is the concept of formalizing the scientific approach to the practice of medicine for identification of “evidence” to support our clinical decisions. It requires an understanding of critical appraisal and the basic epidemiologic principles of study design, point estimates, relative risk, odds ratios, confidence intervals, bias, and confounding. By using this information, clinicians can categorize evidence, assess causality, and make evidence-based recommendations. Evidence-based medicine allows analysis of complicated material so that we can make the best possible clinical decisions for the populations we serve.(Am J Obstet Gynecol 2001;185:275–78.)

Section snippets

Critical appraisal

Grimes5 suggests that understanding EBM is understanding critical appraisal, which is defined as the systematic gathering and synthesizing of the best available evidence using the acumen of the clinician.5, 6 Clinicians can thus understand study limitations, recognize bias, extract information, and reach appropriate conclusions. Critical appraisal is based on well-established epidemiologic principles. Although they are not the only kind of study used in critical appraisal, clinicians rely on

Epidemiology

The narrow definition of epidemiology is the study of the distribution of diseases in a population. The broader definition—an inductive science of biologic inferences derived from observations—is more appropriate when relating EBM principles to clinical decision making. The goal of epidemiology is to identify exposures, make observations, and estimate outcome relationships in order to determine if a causal relationship exists. Classically these studies have been quite good at the “who,” the

The basis for the evidence

Analytic and descriptive are the two fundamental types of epidemiologic studies. Analytic studies attempt to associate a given exposure to (or dissociate from) a risk of a particular disease or outcome. In the hierarchy of analytical studies, experimental designs carry more weight than observation ones for EBM purposes. Experimental design, such as the RCT, is considered the gold standard because the investigator controls the exposure of the factor of interest. However, because of the inherent

Categorizing evidence

Epidemiologic principles allow us to categorize evidence. Several groups have made significant contributions to developing strategies for categorizing the quality of evidence, such as the Canadian Task Force on the Periodic Examination,12 the U.S. Preventive Services Task Force13 and the Cochrane Collaboration.2 Level I evidence is obtained from at least one properly controlled randomized trial and is considered the gold standard of evidence.3 Level II-1 evidence is derived from controlled

Assessing causality

The assessment of cause and effect starts with determining the statistical relationships. If there is no association (RR close to 1.0) there is no cause and effect. If there is an association, assessing the magnitude and precision of the relationship is the next step. If there is an important, precise relative risk, then the criteria proposed by Sir Austin Bradford-Hill should be used to determine causality.15 The Bradford-Hill Criteria for Causation, first published in 1968, provide guidelines

Comment

Critics of EBM claim that all clinical trials have inclusion and exclusion criteria that make their subjects not “real world.” They also point out that subjects and investigators are human and, if given the opportunity, may unconsciously subvert randomization; that not all trials are correctly analyzed or reported; and that epidemiologic risk may not equate to individual risk. However, EBM, while never perfect, allows clinicians to synthesize complicated material so that the best-informed

References (16)

There are more references available in the full text version of this article.

Cited by (17)

  • Efficacy of Varicocelectomy in Improving Semen Parameters: New Meta-analytical Approach

    2007, Urology
    Citation Excerpt :

    In the present meta-analysis, we analyzed randomized controlled trials (RCTs) and observational studies. Although RCTs remain the reference standard, some investigators have recommended inclusion of case-control and observational studies in a meta-analysis, as long as the studies have been carefully evaluated for potential bias.16,17 The studies were selected by performing an extensive electronic and hand search using BIOSIS, EMBASE, and Medline from 1985 to the present with the help of a professional librarian.

  • Reassessing the value of varicocelectomy as a treatment for male subfertility with a new meta-analysis

    2007, Fertility and Sterility
    Citation Excerpt :

    Although this approach may be controversial to some, we believe that it is sound for the current subject matter primarily because of the lack of reliable randomized controlled trial data. Several studies suggest that randomized controlled trials provide the highest level of evidence for causation, but they are known to be costly and difficult to complete, particularly without experiencing a significant number of dropouts after randomization (21, 27, 28). Furthermore, in some instances, the ethics of the randomized controlled trials may be viewed as borderline because the randomization and informed consent may not reach international standards, and in most cases they are done without peer review (40).

  • For systematic hysterectomy during genital prolapse surgery

    2006, Gynecologie Obstetrique et Fertilite
View all citing articles on Scopus

Reprint requests: J. Kell Williams, MD, Department of Obstetrics and Gynecology, University of South Florida College of Medicine, 4 Columbia Drive, Suite 529, Tampa, FL 33606. E-mail: [email protected]

View full text