Acquire evidence Flashcards
What is internal validity?
Can the evidence given be trusted? i.e are the study results accurate and reliable?
An internally valid study is one which measured what it set out to.
What is external validity?
To what extent do the results of a study apply to people outside the study population?
What should the risk of bias be like in a study with high internal validity?
Very low
What are confounding and bias? What other factor can influence study outcome?
They describe factors that may influence or cause a study’s outcome.
Study outcome may be influenced by chance.
What are the types of bias that trialists refer to when assessing risk of bias for RCTs?
The cochrane tool is used which assesses:
Selection bias
Performance bias
Detection bias
Attrition bias
Reporting bias
What different terms are used by epidemiologists to describe bias?
Confounding
Selection bias
Information bias
Which tool is most reliable when assessing bias?
The cochrane tool
What is a confounding factor?
Objective of an experiment: A causes B.
Other causes can cause B that are not A and they are confounding factors..
How are confounders minimised or eliminated?
Using several strategies such as randomisation
What strategies can be used to reduce confounder effect on observational studies?
Restriction
Matching
Stratification
Statistical Techniques
How does randomisation reduce confounder effects?
Random allocation of participants between groups should result in a random and even distribution of confounders between treatment and control groups.
It does not require knowledge of possible confounders making it a good control for both known and unknown confounding factors.
What is restriction?
Exclusion or specification which involves screening out participants if they possess a confounder.
What are the disadvantages to restriction?
Will slow recruitment
Increases internal validity while decreasing external validity/applicability.
What is matching and what are the cons to doing it?
Matching a participant that has a confounder with someone with the same confounder in the control group.
The problem with this is it can be onerous if there are several confounders. The study can no longer examine the effect of a matched variable as a result of this.
What is stratification?
After study completion a restriction approach is taken and then comparisons can be made between the 2 groups to confirm the effect of a confounder. This strategy facilitates recruitment and enables assessment of the influence of a confounder.
What is bias? What does it do to results?
Systematic error in collecting and interpreting data. Can result in both over and underestimating a study effect.
What does bias do to validity of a study?
Bias undermines internal validity of a study.
What is selection bias?
Systematic differences between baseline characteristics of the groups that are compared
What is performance bias?
Systematic differences between groups in the care that is provided or in exposure to factors other than the interventions of interest. (one group gets looked after or receives more attention than the other)
What is detection bias?
Systematic differences between groups in how outcomes are determined.
Differences in how outcomes are measured between groups can be a source of detection bias.
What is attrition bias?
Systematic differences between groups in withdrawals from a study
What is reporting/publication bias?
Systematic differences between reported and unreported findings.
*Studies with positive outcomes tend to be preferentially published and published faster than negative findings
How can attrition bias be avoided?
Over-recruiting
When there is knowledge of the bias, post-hoc analysis strategies can be used.
When does attrition bias occur? (what kind of attrition?)
When attrition is NOT random. It must also be at a rate >20%
Where is publication bias problematic?
With the production of meta-analyses from the present publications.
Why does publication bias happen?
Highly competitive research environment
Career promotion concerns
Competition for funding.
What is the funnel plot?
A scatter plot plotting the result (Often standard error of the effect) and precision (odds ratios) of multiple studies.
What kind of funnel plot indicates bias?
Asymmetry. The plot should resemble a pyramid.
