HSPH core stats overview Flashcards
What is the standard deviation?
Average distance from the mean
What is the standard error?
SD of a sample population
Calculate SE
SD/square root of sample size
Calculate CI
Mean +/- 1.96xSE
Calculate relative risk
RR = a/b
a = risk of x in the treatment group b = risk of x in the control group
RR of treatment that has no effect is?
RR = 1
What is absolute risk?
Risk in the placebo group - risk in the treatment group
Absolute risk is used to calculate what?
Calculate the risk in a randomised controlled trial
Calculate absolute excess risk (AER)
Risk in exposed group - risk in unexposed group
Significance of AER is?
?
AER is used in what study type?
Cohort trial
Calculate the attributable population
Incidence in the population attributable to exposure/incidence in the population
1 (relative risk - 1) / 1+p (relative risk -1)
p = proportion exposed e.g. 20% = 0.2
Absolute risk is used in what study type?
Randomised controlled trial
Attributable population is used in what study type?
Cohort study
Calculate the odds ratio
a/b
a = odds exposure in cases
b = odds exposure in controls
When is the OR used?
In case control studies in place of the RR
Have crude and adjusted
What are cluster randomised trials?
Randomise pre-existing groups into one of two treatments e.g. schools, GP surgeries, villages
How do you determine a great enough sample size?
Should have a statistical power of at least 80% - usually of at least 85%
What is a nested case control study?
Identify a cohort and study them throughout the years
Later years - those who e.g. died are the cases and those that did not die are the controls
Can then look back on the original serum samples to compare the cases and controls
What is a cross sectional study?
Snapshot study - measure existing disease and current exposure with no follow up and no questions about history
