Population Science Flashcards
is the 95% confidence interval includes the number 1 what does this mean about the results?
they are not statistically significant. The null hypotheses is rejected.
what is blinding
where the patient is not aware of which treatment they are being given
what are the advantages of blinding
minimises allocation bias, behaviour changes of the patient, measurement bias
what is measurement bias
how assessors may later the way they collect information
what are appropriate losses to follow up
where the patient has to be removed from the trial due to a deterioration in their health
what are unfortunate losses to follow up
where a patient chooses to withdraw from the trial
why may patients not be compliant with their treatment
didn’t understand instructions
feel better already
don’t like the treatment
how can you improve patient compliance
explain instructions fully
directly observed treatment
what is As-Treated analysis
where you only analyse the results from those how fully complied with treatment and discard follow up losses
what is Intention-to-treat analysis
where the analysis of the trial is done according to original allocation regardless of follow up losses and compliance
True or False: As-treated analysis gives a more realistic size of the effect of the treatment
false - its intention-to-treat analysis
why are RCTs used
they remove confounding and allocation bias
whilst also allowing for reproducible, controlled and fair trials
what are RCTs
where 2 identical groups are produced to make a fair comparison when looking at 2 different treatment options
what is a primary outcome measure
the main and preferably only outcome
what is a secondary outcome measure
other outcomes of interest
what is the placebo effect
where patient attitude to their illness or the illness itself may be improved if the patient feels that something is being done about it
what is used to remove the placebo effect
placebos
what is a placebo
an inert substance made to appear identical to a treatment
What are systematic reviews
Where lots of similar studies answering the same question are brought together
What is meta-analysis
Where the studies in a systematic review are pooled together to get one overall result
What is the purpose of meta-analysis
Collecting the study results means you can make more certain statements about the information as it based on a larger group.
Also reduces problems in interpreting the data
How is meta-analysis presented
On a Forrest plot
How are studies on a Forrest plot weighted
Those with more certainty (smaller CIs) and a bigger study size are weighted more heavily
What is the random effect model
A difference in a Forrest plot taking into account variation between studies
How do you interpret Forrest plots
Boxes = odds ratio
Horizontal lines = CI of the study
Diamond = odds ratio of pooled estimate and its width is the CI
What can be used to explain variation between studies
Sub group analysis - this looks at the study characteristics and the patient profiles
What is publication bias
Where studies which are statistically significant are more likely to be published than those which aren’t
What plot is used to determine if publication bias is present
A funnel plot
