HESC 401 Chapter Notes - Chapter 8: Blind Experiment, Endometrial Cancer, Publication Bias
15 Oct 2016
School
Department
Course
Professor
Chapter 8. Further Issues in Randomized Trials
Objectives
To define key concepts of epidemiologic study design in the context of
randomized trials: sample size, type I error, type II error, power,
generalizability (external validity), and internal validity.
To calculate and interpret efficacy in a randomized trial.
To describe the design and results of five important randomized trials.
To define the four major phases of randomized trials that are used by the
Federal Drug Administration for evaluating new drugs in the United States.
To introduce several ethical considerations as they relate to randomized
trials.
To discuss the rationale for requiring the registration of newly initiated
randomized trials.
Sample Size
You need to have a large number test subjects in order to have
meaningful results.
If you have too small a sample size, then the results of your study is
making assumptions regarding the distribution of the general
population.
Interpreting the Results of Randomized Trials
Because the risk of developing a disease can be calculated form the
results of randomized trials, efficacy can also be calculated.
Efficacy: measures how much disease can be prevented as a result of
the intervention method studied in the trial.
o Can be measured as:
((Rate in those who received placebo) – (Rate in those who
received the vaccine))/ (Rate in those who received the
placebo)
find more resources at oneclass.com
find more resources at oneclass.com
2
External validity: occurs within a trial when its results can be said to
apply to an entire population.
o The study population has been carefully examined as being
representative of the general population.
Internal validity: occurs when the findings of the study are applicable
to the population within the study itself.
o This means that biases and other factors have been removed
and results are consistent throughout the study groups.
There have been three major randomized trials in the U.S.
The Hypertension Detection and Follow-up Program
The Multiple Risk Factor Intervention Trial
Study of Breast Cancer Prevention Using Tamoxifen
The Hypertension Detection and Follow Up Program
Study onduted y the Veteran’s Administration that demonstrated
that treating people who have large increases in blood pressure
reduce their mortality.
The eligible population for the study was divided into two groups and
their mortalities were studied over a 5 year period:
o Referred care: subjects were referred back to their physicians.
o Stepped care: treatment according to a defined protocol in but
changed when a decrease in blood pressure did not occur after a
period of time.
These patients had a lower mortality between the two
study groups.
Multiple Risk Factor Intervention Trial
Randomized trial that sought out to determine whether mortality
from myocardial infarction could be reduced.
The eligible population for the study was divided into two groups:
o Special intervention group: received stepped care.
o Usual Care group:: received usual care provided in the
community.
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
To define key concepts of epidemiologic study design in the context of randomized trials: sample size, type i error, type ii error, power, generalizability (external validity), and internal validity. To calculate and interpret efficacy in a randomized trial. To describe the design and results of five important randomized trials. To define the four major phases of randomized trials that are used by the. Federal drug administration for evaluating new drugs in the united states. To introduce several ethical considerations as they relate to randomized trials. To discuss the rationale for requiring the registration of newly initiated randomized trials. You need to have a large number test subjects in order to have meaningful results. If you have too small a sample size, then the results of your study is making assumptions regarding the distribution of the general population. Because the risk of developing a disease can be calculated form the results of randomized trials, efficacy can also be calculated.
