EPIDEMIOLOGY and BIOSTATISTICS: SKILLS FOR CRITICAL
READING OF THE MEDICAL LITERATURE
Course Director: Judy Savageau, MPH
Department of Family Medicine and Community Health
The primary goal of this course is to provide medical students with a core set of epidemiologic and biostatistical concepts and skills required to critically evaluate research reports in the medical literature.
In this course, students are expected to learn the concepts and skills outlined below and be able to apply these skills to solve epidemiologic problems and interpret research reports in the medical literature.
This short course is predominantly taught in small group workshop settings, with a few large group lectures. These workshops are designed to give students direct experience in using the skills and concepts covered in the course. Workshops usually begin with a 15-20 minute lecture/discussion covering the main concepts in the lecture notes related to the topic for the workshop (e.g., bias, study design). Following the discussion of concepts, students work on the exercises in the workshop manual, often in small groups, and discuss the exercises under the direction of a faculty facilitator.
COURSE GOALS AND OBJECTIVES
- To assist students in developing a working vocabulary of relevant technical terms required for intelligent reading of the medical literature.
- To assist students in developing the analytic skills required to accurately and critically evaluate scientific clinical medical evidence about treatment, prevention, prognosis, and etiology as presented in original research reports, in reviews and practice guidelines, and in debates of controversial issues within the medical community.
- To develop in students an awareness of contemporary approaches and issues in clinical research in medicine.
By the end of the course, students should be able to:
A. Measures of Disease Occurrence and Association
- crude mortality rate
- specific mortality rate (age, sex, race, and cause)
- case fatality rate
- Cite one example of the correct use of each rate listed above, and interpret statements containing them.
- State the reasons for adjustment of rates and interpret statements containing adjusted rates.
- Define incidence and prevalence; state the relationship between them. Name the factors that may cause variation in each measurement. Give the uses of each rate and major sources of data for constructing them.
- Define relative risk and attributable risk. Interpret statements that employ these terms. Compute relative and attributable risk from the data in a two by two table.
- Define odds ratio and interpret this measure when used in a cross-sectional or case-control study.
B. Design of Epidemiologic Studies
- Distinguish between experimental and non-experimental studies as well as between descriptive and analytical studies.
- Describe the following types of epidemiologic studies and discuss the appropriate use of each design and the strengths and weaknesses of each:
- descriptive (case reports/series and descriptive cohort studies)
- cohort (prospective and historical prospective)
- clinical trial
C. Bias and Cause in Epidemiologic Studies
- Give examples of the major ways in which biasor nonrandom error can lead to wrong conclusions (inference) in studies or clinical decisions.
- Give a working definition of confounding, cite an example, and describe the common ways for dealing with confounding in research studies.
- Distinguish between association and causation and discuss how causal inferences are made in medicine and public health.
- State the purpose of a frequency and relative frequency distribution and a frequency histogram in describing a set of data.
- Define mean, median, mode, and percentile, and describe the features of a distribution that each characterizes.
- Contrast the features of a normal (Gaussian) distribution with those of a skewed distribution.
- Describe the basic concepts involved in making statistical inferences about results obtained from study samples.
- Describe the basic concepts involved in deriving p values and in developing confidence intervals.
- Give examples of the major ways in which chanceor random error can lead to wrong conclusions (inference) in studies or clinical decisions.
- Interpret statements of statistical significance with regard to comparisons of means and frequencies and explain what is meant by a statement such as "p<0.05."
- Describe the limitations of p values as a way of expressing the role of chance and the advantages of confidence intervals.