Flashcards in Epidemiology Deck (35):
refers to existing cases
refers to NEW cases over an unaffected population
how to calculate risk?
risk= new cases/population at risk
how to calculate rate?
rate= new cases in time period/ total follow up
what is hazard?
instantaneous rate for longitudinal studies
how to calculate relative risk?
risk exposed/ risk Unexposed (re/ru)
how to calculate attributable risk?
risk exposed - risk unexposed (re-ru)
how to calculate attributable risk %
(attributable risk/ risk exposed) x 100
what do we mean by non longitudinal and longitudinal studies?
longitudinal= follow up
non longitudinal= no follow up
what types of studies are found in longitudinal studies
what types of studies are found in non longitudinal studies?
describe a CASE control trial
Odds ratio calculated (likelihood of getting disease)
how do accurate odds ratio?
what are the two types of bias?
how to prevent selection bias?
careful recruitment, minimise loss to follow up
how to prevent information bias?
what is a confounder?
a variable that influences the relationship between Exposure and Outcome
what is the gold standard sort of trial that ix evidence of causality?
what are some key considerations of clinical trials?
intention to treat analysis
number needed to treat
how to calculate the hazard ratio HR?
hazard (intervention) : hazard (control)
what is the number needed to treat?
the no of ppl that are needed to undergo the intervention in order to prevent outcome in one
what does it mean when HR (hazard ratio) is 0.8?
the probability that the intervention will reduce outcome is 20%
why is NNT important in a clinical trial?
demonstrates the efficacy of the intervention
how to calculate NNT?
1/ absolute risk or rate reduction
what do we mean by absolute reduction?
difference between rate of outcome in control group and rate of outcome in intervention group
what type of study is considered the best level of evidence
randomised control studies/clinical trials
how do we assess internal validity?
what does a small confidence interval suggest?
narrow= precise or smaller sample size
what does are large confidence interval suggest?
wide= less precise or larger sample size
how to calculate SENSITIVITY
true positive/ (true positive + false negative)
TP/ (TP + FN)
how to calculate SPECIFICITY
true negative/ (true negative + false positive)
TN/ (TN + FP)
how to calculate PPV (positive predictive value)
true positive/ (true positive + false positive)
TP/ (TP+ FP)
how to calculate NPV (negative predictive value)
true negative/ (true negative + false negative)
TN/ (TN + FN)
how to calculate likelihood ratio of +ve test?
sensitivity/ (1- specificity)