EVALUATION OF AN EPIDEMIOLOGIC STUDY

1
EVALUATION OF AN EPIDEMIOLOGIC STUDY

• Melissa Fleschler/Sharon Cooper
• July 19, 2001

2
The basic structure of an article

• Abstract/Summary

• Overview or summary
• Short description of sample
• Highlights of results
• General statement of significance

3
Introduction

• Background information
• history, pathophysiology, descriptive
  epidemiology of problem
• Review of work of others
• Rationale for present study
• Study Questions/Hypothesis

4
Methods/Materials and Methods

• Study design
• Subject selection procedures
• Description of Sample
• Description of Independent and Dependent
  Variables
• Methods of measurement
• Description of analytic techniques
• Quality control procedures
• Specific analyses to be performed

5
Results

• What happened?
• Descriptive Statistics
• Univariate, Multivariate analyses, etc.
• Graphicstables, charts, figuresthat summarize
  findings

6
Discussion/Comment/Conclusions

• Meaning, significance of work
• Comparison of work with others
• Critique of study
• discussion of limitations (e.g., type of biases
  present) as well as strengths
• Disclaimers, apologies, speculation, fantasy

7
References/Bibliography

• Evidence that work of others has been considered
• Leads to further exploration of the subject

Ref Gelbach, 1993
8
Paper Critique
9
THE QUESTION

• What is the research objective being answered?
  What is the hypothesis being tested?
• Is it the original objective of the authors or
  secondary to their original research?
• If it is secondary, was the reported research
  question formulated after the data were analyzed
  (e.g., data driven)?

10
THE DESIGN

• Is the study design that was used clearly
  definedwhat is it? Is this study design
  appropriate?
• Identify the target population, study population
  and examined population.
• What was the response rate?

11
THE CONDUCT

• Were the study variables identified and defined?
  (e.g., what were the exposure and outcome
  variables?)
• How were the major study variables measured?

12
THE ANALYSIS

• What measure(s) of epidemiologic association was
  calculated?
• Is the analysis appropriate for the study design?
• If necessary, were potential confounders
  controlled for in the analysis?
• Were confidence intervals calculated?
• Were p values reported?
• Are the findings internally consistent (i.e., do
  tables add up and are different tables
  numerically consistent?)

13
THE RESULTS

• What were the major results as presented by the
  authors?
• How large is the observed association?
• Is there evidence of a dose-response relationship?

14
VALIDITY

• Internal validitywas the potential presence of
  bias addressed in either the design of the study
  or in the analysis?
• Selection bias
• Information bias
• Confounding
• In what direction would each potential bias
  influence the results?
• External validityto which population(s) can the
  results be generalized?

15
THE INTERPRETATION

• Are the findings consistent with other studies?
• What was the author(s) interpretation?
• Do you agree with this interpretation?
• Did the authors clearly state the limitations of
  their research?
• Do you believe there are any other limitations or
  biases inherent in the study that affect the
  author(s) interpretation?