Lecture 6, part 2 Flashcards
What should be considered when choosing a pop?
Potential for meeting study objectives
Comparability
Method of recruitment
Eligibility criteria
Sample size
Generalizability
Consent
Comparability details
Select participants to reduce heterogeneity
What are some methods of recruitment?
Pt records
Ads
Clinician referrals
Details about eligibility criteria
Well-defined and appropriate criteria based on the research question will help control/reduce errors/bias
Definition of generalizability
What reference pop can the results be applied to?
Sample size guiding principles
Rarer outcomes require larger sample sizes
Need to start with some idea of what differences might exist in order to have a good chance of detecting differences of clinical importance
What samples are required in terms of detecting differences?
Detecting small differences requires larger sample
Detecting large difference requires smaller sample
What is the twofold goal of randomization?
To create comparability with respect to other potentially known and unknown confounding variables
To ensure that tx allocation/assignment is not influenced by investigators
-Prevents potential biases on the part of the investigators from influencing the assignment of participants to different tx groups
Essential components of randomization
Tx assignment must be made free of influence from both the study participant and study personnel
Future assignments must not be able to be predicted from past assignments
The order of the allocations must be reproducible and documentation of the method provided
What can help randomization work?
Before randomization: prevention through blocking or stratification
After randomization: apply analytic strategies
Blocking definition
Randomization is conducted in groups of certain size to ensure that equal numbers of pts are assigned to each of the two groups
Stratification definition
Used to ensure that key confounding variables are equally distributed between the tx and comparison groups
Individuals first stratified according to confounding characteristics, then randomly assigned to tx or comparison group
Aspects of measuring baseline variables
Characterize/describe the study pop
Measure variables that are predictors of the outcome (potential confounders)
Ensure that there is balance (did randomization work?)
Components of applying the intervention
Develop intervention
Implement intervention
Monitor intervention
Steps in developing the intervention
Develop materials and protocols
Train staff (standardize training, use training materials)
